The Homophonic Forms of Musical Composition: An Exhaustive Treatise on the ...

Google

This is a digital copy of a book that was preserved for generations on library shelves before it was carefully scanned by Google as part of a project

to make the world's books discoverable online.

It has survived long enough for the copyright to expire and the book to enter the public domain. A public domain book is one that was never subject

to copyright or whose legal copyright term has expired. Whether a book is in the public domain may vary country to country. Public domain books

are our gateways to the past, representing a wealth of history, culture and knowledge that's often difficult to discover.

Marks, notations and other maiginalia present in the original volume will appear in this file - a reminder of this book's long journey from the

publisher to a library and finally to you.

Usage guidelines

Google is proud to partner with libraries to digitize public domain materials and make them widely accessible. Public domain books belong to the
public and we are merely their custodians. Nevertheless, this work is expensive, so in order to keep providing tliis resource, we liave taken steps to
prevent abuse by commercial parties, including placing technical restrictions on automated querying.
We also ask that you:

+ Make non-commercial use of the files We designed Google Book Search for use by individuals, and we request that you use these files for
personal, non-commercial purposes.

+ Refrain fivm automated querying Do not send automated queries of any sort to Google's system: If you are conducting research on machine
translation, optical character recognition or other areas where access to a large amount of text is helpful, please contact us. We encourage the
use of public domain materials for these purposes and may be able to help.

+ Maintain attributionTht GoogXt "watermark" you see on each file is essential for in forming people about this project and helping them find
additional materials through Google Book Search. Please do not remove it.

+ Keep it legal Whatever your use, remember that you are responsible for ensuring that what you are doing is legal. Do not assume that just
because we believe a book is in the public domain for users in the United States, that the work is also in the public domain for users in other
countries. Whether a book is still in copyright varies from country to country, and we can't offer guidance on whether any specific use of
any specific book is allowed. Please do not assume that a book's appearance in Google Book Search means it can be used in any manner
anywhere in the world. Copyright infringement liabili^ can be quite severe.

About Google Book Search

Google's mission is to organize the world's information and to make it universally accessible and useful. Google Book Search helps readers
discover the world's books while helping authors and publishers reach new audiences. You can search through the full text of this book on the web

at |http: //books .google .com/I

I

Prof. k. A. Stanley

THE

HOMOPHONIC FORMS

OF

MUSICAL COMPOSITION

AN EXHAUSTIVE TREATISE ON THE STRUCTURE

AND DEVELOPMENT OF MUSICAL FORMS,

FROM THE SIMPLE PHRASE TO THE

SONG-FORM WITH " TRIO "

FOR THE USB OF

(Penetal anb Spectal Stubents of Ausfcal Structure

A *

PERCY GOETSCHIUS, Mus. Doc.

{Royal Wurtemberg Professor)

author of

"Thb Material used in Musical Composition"
"The Theory and Practice of Tone-Relations"
"Models of the Principal Musical Forms," etc.

it

j New York :

j G. SCHIRMER

1898

-5-2'

Copyright, 1898, by G. SCHIRMER

• » • • .

lano

BURR PRINTING HOUSE*
NEW YORK.

}

I

\

To

ffbvB. I). f>* a* aseacb

AS A CORDIAL TRIBUTE

TO AN EMINENT WOMAN-COM POSER

OF

AMERICA

\

i

■fl^QylU^ ^

(.

PREFACE,

I.

This book undertakes no more than the systematic enumeration
and exhaustive explanation of all the formal designs and methods
of structural treatment in the homophonic domain of musical
composition, as revealed in classical or standard writings. The
student who aims to acquire the Science of composition, is expected
to Imitate these designs and methods, and to look for additional
illustrations and confirmations in general musical literature. This
will develop skill and facility, will induce correct habits of musical
thought, will enrich the mind with a fund of resources, and
stimulate the imagination to increased responsiveness and activity.

But, further than this, the book lays no claim to furnishing
clues to the subtle Art of composition. In converting his theoreti-
cal knowledge into successful practice, — into a means to an end, —
the student can appeal to no other authority than that of his own
fancy, good taste, and natural or acquired judgment.

II.

The examples given for reference must be inspected, — if not
totally, at least in great part. No student should hope to be
entirely successful and efficient as a composer without possessing
quite extensive acquaintance with the products and processes of
successful writers. Therefore, the pupil must regard the conscien-
tious examination of these carefully selected quotations, as a very
significant and distinctly essential part of his study. This applies
more especially to the works cited in Divisions I, II and III.

Some of the works, — those to which constant or frequent
reference is made, — he must endeavor to own ; for example :

The Pianoforte Sonatas of Beethoven, Haydn, Mozart and
; Schubert ;

^ The Symphonies (arr. for 2 hands) of Beethoven and Haydn ;

The Bagatelles of Beethoven ;

t.

The ** Songs without Words " of Mendelssohn ;
The Mazurkas, Nocturnes and Preludes of Chopin ;
The Pfte. works of Schumann (op. 12, 15, 68, 82, 99, 124),
and of Brahms (op. 10, 76, 79, 116, 117, 118, 119) ;

And some of the Songs of Schubert and Schumann.

III.

The author hopes and expects that the book will prove quite
as necessary and useful to the general music-student^ as to the
prospective composer ; a knowledge of homophonic musical struc-
ture being, undeniably, of equal importance to all musical artists,
reproductive as well as productive.

The general student, while studying and analyzing with the
same thoroughness as the special student of composition, will
simply omit all the prescribed Exercises.

Percy Goetschius, Mus. Doc.

Boston, Mass., December, 1897.

TABLE OF CONTENTS.

(Figures in parentheses refer to paragraphs.)

Introductory :

The requisites of musical composition

The harmonic fundament

Basis of chord-succession

Ornamentation of chords

Figuration of chords

The Divisions of musical form

PAGE

I

2
2

4
5

DIVISION I.

Chap. I. The Phrase :

Definition (i)

Beginning and ending (2) .

Perfect cadence (3) .

Modification of perfect cadence (4)

Harmonic aspect of phrase (5)

Approach to the perf. cadence (6, 7)

Melodic aspect of phrase (8)

Division of phrase-melody into members (9)

Means of indicating melodic joints (10)

Syntax of phrase (11)

Relations between melodic members (12)

Exercise 1.

7

7

8
8

9
10

12

16
16

Chap. II. The harmonic kc^ipmbnt of

THE PHRASB-HBLODY :

Styles of accompaniment (13) . . 19
Distinction of one-, two-, three-, and four-
voiced styles (14) . . . • ^9
Irregular harmonic bulk (15) . -23
JSxercise 2.

TABLE OF CONTENTS.

Chap. III. The development or exten-

sion OF THE PHRASE :

Processes of composition (i6)
Means employed in developing phrase (17)
Repetition of entire phrase (18)
Concealing perfect cadence (18)
Modifications of repetition (19) :

Embellishment of melody (i^a) .

Change of harmony and modulation (193)

Change of register (19c) .

Change of style (igd)

Complete changes in course (19^) .
Interlude before repetition (20)
Conditions upon which ** repetition" de

pends (21)
Series of repetitions (22)
Object of repetitions (23)
Exercise 3.
Extensions at the end of phrase (24) :

Repetition of second half (25a) .

Repetition or sequence of last member
(253) ....

Repetition of cadence-group (26a)

Repetition of two cadence-chords (263)

Reiteration of final tonic chord (26c)

Plagal extension {26d)
Object of extensions at end (27)
JSxerctses 4, and 5.

Extensions at the beginning of phrase (28)
Extensions in course of phrase (29) :

Repetition of early member (29a)

Sequence of early member (29^) .

Expansion of prominent tone or chord
(29c) ....

Importance of " expansion "

New cadence-member (29^)
Extensions in course, when and where

appropriate (30) . . . .

Exercises 6 and 7,

PAGE

36

36

38
38
39

30
31
32

33
33
33

34

36

37
38
40

41
41

42

45

45
46

48

52
53

54

TABLE OF CONTENTS.

XI

Chap. IV.

Chain-phrase (31) .
Melody-expansion (32)
Irregular phrase-formation (33)
Miscellaneous examples of phrase-extension
Exercise 8.

PAGE

55

57

59
61

Chap. V. The Period-form :

Definition (34)

Antecedent phrase (35)

Semicadences (36) .

Character of antecedent phrase (37)

Consequent phrase (38)

Construction of consequent phrase (39)
Parallel construction (39<aj)
Opposite construction (39^)
Contrasting construction (39c)

Variety and Unity (40)

Exercise 9.

62
62
62

63
64

64

64

67

69

70

Chap. VI. Extensions of Period-form :

Repetition of entire period (41) . . 72

Repetition of consequent-phrase (42) . 73

Repetition of antecedent-phrase, or both

phrases (43) . . . -75

Exercises 10 and 11.

Extensions at beginning of period (44) . 77

Introductory phrase . . . -77

*' Prelude" (45) . . . .78

Introduction to consequent-phrase (46) . 78
Extensions at end of period (47) . . 79

Extension at end of antecedent-phrase (48) 81
Extensions in course of period (49) . . 82

Chain-phrase formation of either phrase (50) 83
Codetta (51) . . . . .84

**Postlude" (52) . . . .86

Natural location of extensions (53) . . 87

Miscellaneous examples of period-extension 87
Exercises 12 and 13,

xu

TABLE OF CONTENTS.

re

Chap. VII. Group-formations :

Period with consequent-group (54)
Distinction between *' repetition " and

production" (55):

Essentially different cadences

Sequential reproduction .

Essential modifications
Period with antecedent-group (56)
Group of phrases, similar (57)
Group of dissimilar phrases (58)
Period-group (59)
Elision (60) .
Exercises H and 15,

Chap. VIII. The Double period

Definition (61)

Inter-relations (62) .

Cadence-conditions (63)

Parallel construction (64) .

Indefinite transitional grades, between single

and double period (65)
Contrasting construction (66)
Extensions of double period (67)
Addition of extraneous members (68a)
Miscellaneous examples of double-period

extensions .
Quadruple period (68^)
Exercises 16 and 17 »

PAOE

88

88
88
90

93

93
95
97
97

lOI
lOI

102
102

104
106
107
no

»

no
III

DIVISION 2.

The Song-forms or Part-forms.

Comparative definition of "Phrase,"

** Part," and *' Song-form " (69) .
Absolute definition of ** Part " (70)

Chap. IX. The Two-part Song-form :

Definition (71)

Details of First Part : form (72a) .

Cadence (72^) .

Modulation (72^) . . . .

"3
114

"5
116

TABLE OF CONTENTS. XIU

PAGE

Details of Second Part : form (73«). . ii6

Character (73^) . . . .116

Cadence (73c) . . . • ^^7

Two-part song-form, primary design (74) . 117

Coincidence of endings of two parts (75) . 120

Exercise 18.

Diminutive two-part song-form (76) . 121

Perplexing external resemblance between

different formal designs (77) . .123

Exercise 19.

Chap. X. The fully developed Two-
part Song-form :

Definition (78) . . . '125

Extensions: repetitions (79a) . .125

Extraneous adjuncts (79^) • • ^^5

Proportion to length of form (79^) • 126

Large two-part form, as type of Sonatina-
form (80) ..... 128

Exercise 20.

Chap. XI. The Three-part Song-form :

Ruling principle of all tripartite forms (81 a) 129
Distinction between ''recurrence" and

" repetition " (8i3) . . . 130

Influence upon Part II of tripartite form

(8ie:) ..... 130

The three-part period (82) . . -131

Genuine species (830?) . . • ^3^

Irregular species (83^) . . • ^33

Exercise 21.

Incipient grade of three-part song- form (84a) 135
Details of structure (84^) . . • ^35

Demonstration of this classification (foot-
note) ..... 138
Resemblance between three-part period and

incipient three-part song-form (84c) . 139

Exercise

V

XIV

TABLE OF CONTENTS.

Chap. XII. The ordinary complete Three-
part Song-form :

PAGE

Definition (85) ^ . . -139
Details of Parti : design (86«), length (866) 140
Details of Part II — ^Thematic conditions :

Total agreement with Part I (87a) . 140

Derived from fragments of Part I (87^) . 141

Opposite construction (Sjc) . . 144

General formative agreement (87^) . 144

New melodic contents (87^) . . 146

Tonality of Part II (88) . . .147

Structural design of Part II (89a) . • ^49

Sectional form of Part II (89^) . • 150

Cadence of Part II :

On dominant {90a) . . '152
On other chords (90^) . .154
The re-transition (90c) . . • ^57
Details of Part III (91) . .162
Extraneous members: codetta or coda, in-
troduction, interlude (92). . .164
Exercises 2S, 24, 25.

Chap. XIII. Additional details of the
Song-forms :

Irregular cadence-conditions :

Imperfect cadence at end of section (93^) 166
Perfect cadence in course of section (93^) 168

Influence of thematic idea upon formal de-
sign (93^:) . . . . .170

Modulation: transient (94^) .' .170

Complete (94^) . . .. • ^7'

Two general rules (94c) . . • ^7^

General modulatory design (94^/). . 172

Dynamic design (95^) . . .172

Other expression-marks (95^) . • ^73

Contrast (96) . . . -173

Style : time (970^) . . . • ^ 74

Tempo (97^), mode (97c), rhythm (^jd) . 175
Emotional elements (97^) . . ••^77

TABLE OF CONTENTS. XV

PAGE

Coda and codetta : distinction between (98) 178
Object and design of coda and codetta

(9805) . . . . .178

Derivation of coda and codetta (98^) . 179
Location of codetta (98c) . . .180

Exercise 26.

Chap. XIV. The incomplete Three-part
Song-form :

Definition (99) . .181

Distinction between incipient grade and in-
complete form (99a) . . .181

Part III a slightly contracted version of
Part I (99^) . . 183

Part I briefly represented at beginning of
Part III (99c) . , . . -183

Exercise 27.

Augmented two-part Song-form (100) . 186

Exercise 28.

Chap. XV. Fully developed Three-part

SONG-FORM :

Definition (loi) . . . .188

Four stages of progressive development :
Stage I, literal, or unessentially modified

recurrence {102a) . . .188

Stage 2, recurrence of portion, extended

(io23) ..... 188
Stage 3, recurrence containing new ma-
terial (io2c) .... 190
Stage 4, recurrence containing member
from Part II (102^) . . . 191

Corroboration (103) . . . '193

Exercises 29, 30, 31,
Large phrase-group {104) . . * ^9S

Chap. XVI. The evolution of the Five-
PART Song-form r
Repetition of the divisions (105a) . .196

Second Part not to be repeated alone (105^) 197
Exercise 32,

XVI TABLE OF CONTENTS.

PAGE

Modified repetition of second division :
Stage I, repetition of Parts II and III

with unessential changes (106^5) . 198

Stage 2, transposed recurrence of Part II

(1066) ..... 198
Stage 3, Part IV a transposed and altered

recurrence of Part II [io6c) . . 200

.Stage 4, Part IV a reconstructed version

of Part II (loSd) . . . 201

Stage 5, Part IV new (106^) . .201

Treatment of Part V (107) . . . 202

Old-fashioned rondeau (io8a) . . 203

Seven-part form (108^) . . . 203

' Exercise SS,

Chap. XVII. Irregular Part-forms :

Definition (109) .... 204

Transposed 3rd Part (no) . . . 205

Group of Parts, incipient stage (m) . 206

Distinct 2nd Part (11 2fl5) . . . 207

Repetition of 2nd Part alone {112b) . 207

Sequential or transposed reproduction of

Parts (113) . . . . 207

Group of Parts, developed (114) . . 208

Extended (115) .... 208

Exercise 34..

DIVISION 3.

Compound Song-forms.

Definition (ii6) . . . . 210

Chap. XVIII. Song-form with one
*'Trio":

Definition (117) ... . . 210

Details of principal song (iiSa, b) . . 211
Transition (ii8c) .... 211
Details of subordinate song or *' Trio" :

Character (119) . . . .212

Time (119a), key (1193) . • •213

Tempo (119c), design (119^) . . 214

TABLE OF CONTENTS.

Re-transition (119^) . . . .

Da capo (120) . . . .

Coda (121) .
Miscellaneous examples

Relation of five-part form to ' ' song with
Trio" (122) . . . .

Exercise 35.

Chap. XIX. Extension of ** song with
Trio " :

Repetition of *' Trio" and da capo (123) .
Song-form with two "Trios" (124)
Group of song-forms (125) .
Exercise 36.

xvii

PAGE

215
215

217
217

218

219
219
220

DIVISION 4.

Conventional styles of composition.

Classification (126) . . . .222

Definition of Lyric class (126a) . ,222

Definition of fetude-class (126^) , . 222

Definition of Dance-class (126c) . .223

Distinctions approximate (126^?) . . 223

Sources of information (127) • .223

Chap. XX. The lyric class :

Song, with words. Details: text (i28<3j) . 224
Setting (128^) .... 224
Vocal compass (128c) ; mood of music

{i28d) ..... 225

Notation (128^) .... 225

Form (128/*) .... 226

Accompaniment (128^) . . .227

Instrumental duo (129) . . . 228

Song without words, etc. (130a) . .228

Ballade (130^) . . . .229

**Part-Songs" (131^) . . . 229

Sacred Text (131^) ; ensemble (131c) ;

secular text (1310?) . . . 230

Design (131^) .... 230

XVlll TABLE OF CONTENTS.

Chap. XXI. The i^tude-class : „ „

PAGE

Etude or stud^ (^S^) • • • 231

Toccata, capriccio, etc. (1330:) . . 232

Scherzo (133^) ; formative elements (133^) ;
design (133^) . . . .232

Chap. XXII. The dance-class :

Old dance-species (134) . . . 233

Modern dances (135) . . . 234

March (136) .... 234

Conclusion : criticism (137) . . . 235

INTRODUCTORY.

The four most important requisites of successful musical com-
position are : istly, ample comprehension and command of the
relations and associations of tone; 2ndly, an active and fertile
imagination ; 3rdly, a strong and well-balanced intellect ; and
4thly, the life-breathing attribute of emotional passion.

Of these four, the last-mentioned cannot be acquired ; it musjt
be innately present in the disposition of the individual ; and, con-
sequently, it will not, and cannot, be a subject of consideration in
this treatise upon Composition. On the contrary, the second con-
dition, Imagination (with the no less important faculty of Dis-
crimination), can evidently be cultivated and developed to a large
extent; and one of the chief aims of the following pages will
therefore be, to point out every possible means of arousing and
stimulating this imaginative faculty.

The first-named condition, finally, skilful technical manipula-
tion of the tone-material and its resources, is very largely indeed a
matter of study, and may be acquired by any ordinarily intelligent
student, to an extent proportionate to his application and patience^
Though not the most significant, it is the most indispensable — the
first and fundamental, requirement. Therefore, the mastery of the
principal details of Tone-relation (Harmony) is expected of the
student before he undertakes the study of composition proper.
Adopting the view that much facility (probably the greatest and
best) is to be acquired through the application of this fundamental
knowledge to the construction of musical designs, rather than
through its exercise as an independent object of study, the present
author demands no more preparatory harmonic and contrapuntal
knowledge than will have been acquired by the faithful and
exhaustive study of *' The Material used in Musical Composition " *,

* Latest (4th) ed. G. Schirmer, New York. 1895.

2 INTRODUCTORY.

or **The Theory and Practice of Tone-relations" * (with the
supplementary exercise indicated in its preface), or any other
standard treatise upon Harmony.

The Harmonic Fundament.

The broadest principles of Harmony are herewith recapitulated ;
not for the information of the beginner, who would learn nothing
of value from such a summary review, but in order to afford the
advanced student of Harmony a bird's-eye view of the entire domain
of Tone-association, such as will facilitate his choice and use of the
material in the execution of the given musical designs :

1. The source and material basis of all music, of whatsoever
character and style it maybe, is the Chord— i.e., a structure of
from three to five tones, arranged in contiguous intervals of the
(major or minor) third. Of these chord-structures, there are three
of fundamental rank, erected upon the Tonic, Dominant and Second-
dominant (or Subdominant) of the chosen key. The nature of every
Chord is determined by the key, which latter resolves itself into
a Scale of conjunct tones, constituting the Platform of musical
operations.

2. The movement from one chord into another, whereby all
harmonic vitality is generated, seems to be dictated largely (if not
mainly) by the simple choice between a whole-step progression
and a half-step progression, either of which is invariably
feasible in one or another, or all, of the respective tone-lines (parts
or voices), which describe or delineate the figures of the musical
design.

These chord-changes are Regular when the distinction of
whole-step or of half-step progression is not determined by option,
but by the strict harmonic conditions of the key ; they are Irreg-
ular (not necessarily '* wrong") when the choice is arbitrarily
made, giving rise to more or less frequent changes of key.

This sweeping option (of leading the respective parts upward
or downward a whole step or a half-step) discloses one of the most
comprehensive and inexhaustible sources of harmonic and melodic
motion, including, as it does, every conceivable chord- progression,
regular or irregular. But at the same time it is so vague and
seductive, that the student should not adopt it as a rule for the

♦ New Engl. Conservatory, Boston, Mass. 1894.

INTRODUCTORY.

determination of his harmonic conduct. He must regard it rather
as a result (e. g., of some thematic or formal design) than as a
cause ; and must accept it chiefly as a proof that there can never
be an excuse for harmonic monotony or apathy. It is obvious,
furthermore, that the free exercise of this option can be conceded
only to those who have first become firmly grounded in all the
principles of regular harmonic progression. The following ran-
dom successions illustrate the point in question :

1. ^/s step.

^/l

4^

k ' ' Vi

r^^

ri

I '/,

SL

\

IB.

?E

t

^

±

-lO-

I'll '/, n »2)

is:.

t

■f"

V, ' ' V« ' ♦!) V i/i IV III i/i II ♦2) I 1/, I "S)

*i) Up to this point all the chord-successions are regular, because, in each
separate case, both chords belong to the same key, and are connected
according to the natural law of chord-progression.

*2) These two successions are irregular, because not in conformity with
the rules of diatonic chord-movement.

*3) And this succession (in common with all chromatics) is irregular*

because a mixture of keys takes place.

Wagner.

*4)

-Oh

i

^

Wf

*t3

-^

¥

^^^^^^^^

T

♦4) From the ** Gotterdammerung*'. Some of these chord-successions
can be accounted for in no other way than as a purely optional choice of whole-
or half-step part-progression.

INTRODUCTORY .

Rubinstein.

/I /a '2 /2

— ^ I etc.

^^FT^f r t ^

^/2 V. Vf '/2

*5) This unique cluster of half-step moTements may be demonstrated as
passing-notes upon a regular chord-basis.

3. The tones which form a chord constitute the harmonic
nucleus of the phrase; but, while they may, and frequently do,
appear in their primary condition (see Ex. 10; Ex. 27), they are
often ornamented with those other scale-tones which lie immediately
above or below the individual chord-tones (as Neighboring-notes) ;
sometimes so richly, that a very simple and otherwise perhaps
monotonous measure assumes a more unique and elaborate shape,
and is transformed from the character of rugged or stolid simplicity
into one of greater grace or deeper passion. For illustration (chord-
-tones in large type) :

J. Andante.

2.

*

pS

g-b-d.

f

^J^^L^'^

Haydn.

s^

g-b-d

INTRODUCTORY.

2. Waltz,

^ ^ ^

Chopin.

etc.

'^-

*=*:

^^

5^

I

:?c=3:

3. Allegro.

o-eg.

.c-e-g d-f-a-c.

4. Furthermore, the chord-tones (possibly in connection with
their auxiliary neighbors) are not always presented in one simul-
taneous bulk, but frequently so separated (dispersed or ** broken")
as to produce an animated rhythmic and melodic effect, and to
extend into other, higher or lower, registers than the common and
convenient tone-locality corresponding to the compass of the human
voices. Thus :

1. Allegro. Mendelssohn.

* ^

3.

ijij

m^

\

^,4

i^SL:t*:t-t-Ji.

C-e-g

INTRODUCTORY.

Beethoven,

These fundamental varieties of tone-combination, supplemented
by a large number of inferior distinctions which will be pointed out
in their proper place and order, constitute the Material out of
which the composer develops his artistic creation.

The processes to be pursued in this act : the draughting of the
design, the execution of the details, and the modes of manipulating
the natural material — these it is the province of a treatise upon
*' Musical Form " to expound. ^

THE DIVISIONS OF MUSICAL FORM.

The entire range of musical forms is divided into three grand
classes :

Istly, THE HoMOPHONic FORMS ; embracing the majority of
smaller designs, and a number of the larger ones, and characterized
by the predomination of the simpler styles of harmony, and the
element of single Melody.

2ndly, the Polyphonic forms ; embracing the Invention,
Fugue and Canon, and characterized by the constant association
OF two or more individual Melodies.

Srdlyy THE Higher or Complex forms; embracing the
majority of larger designs, and characterized by the union of
the homophonic and polyphonic principles of treatment, and also
by greater logical continuity and closer affinity between the com-
ponent members of the design. The matter under treatment in the
present volume is that of the first of these classes :

THE HOMOPHONIC FORMS.

THE HOMOPHONIC FORMS.

DIVISION ONE.

THE PHRASE-FORMS.

CHAPTER I.

THE PHRASE.

1 . The Phrase is the structural basis of all musical forms. It
is a series of chords in uninterrupted succession, extending (when
regular) through. ^our ordinary measures in ordinary moderate tempo.

When the tempo is slow (Adagio — Larghetto), or the measures large
(6/8 — 9/8 — 12/8), the Phrase may extend through only two measures; and, in-
versely, in rapid tempi, or when the measures are of a smaller denomination,
the Phrase may contain eight measures. Any other, larger or smaller, number
of measures (3, 5, 6, 7, 9, etc.) must constitute an irregular design, the causes
and purposes of which will be seen in due time (par. 33).

2a. The simple Phrase begins with the tonic chord; or, in

exceptional cases, with the dominant (see Ex. 14, No. i), or

with some other harmony of preliminary effect. The first tone of

the Phrase-melody may appear upon any beat or fraction of the

measure.

In case it begins upon any unaccented beat, or fraction of a beat, it is evident
that the first (apparent) measure will be an incomplete one, and it must be dis-
tinctly understood that such an incomplete measure is merely preliminary and
is never to be counted as first measure^ no matter how little it may lack of being
an entire measure. (This must be borne constantly in mind in counting the
measures in all given references. The first measure in a Phrase, or com-
position of any design, is the first full measure {\ The preliminary beats or
fractions are in reality a borrowed portion of the final measure, and will be
deducted from the latter, as a rule. See Ex. 5 ; Ex. 10 (beginning upon first
accent). Ex. 6; Ex. 7b; Ex. 12 (beginning on last beat of preliminary

THE PERFECT CADENCE.

Par. 4.

measure). Ex. 13, No. 2; Ex. 14, No. 2 (beginning one beat and a half before
first accent) ; Ex. 14, No. i ; Ex. 23, No. 6 (beginning on secondary accent of
preliminary measure); Ex. 7; Ex. 13, No. i; Ex. 22, No. 5 (beginning four
beats before first accent) ; Ex. 41, No. i ; Ex. 45, No. i ; Ex. 66, measure 4.

(b) The Phrase closes with a Cadence; if a simple Phrase,
independent of associates, or the final one of a series of Phrases, it
closes with the so-called perfect cadence ; otherwise with a
semi-cadence (explained in par. 36).

A cadence is an interruption of the harmonic and (particularly) the melodic
current, — a check, or pause, — a point of repose, which marks the conclusion
of some melodic line, — but without affecting the fundamental rhythmic pulse.
Cadences are distinguished, and their employment determined, by the force
and extent of this interruption.

3. Of the perfect cadence there is only one harmonic form,
namely : the Tonic triad, with its root in both outer parts, on an
accented beat, and preceded by a fundamental form of the Dominant
harmony (whose rhythmic location and extent is, however, optional).
Thus, in the skeleton of a 4-measure Phrase :

X. Accented beginning, Perf. Cad. 2. Unacc* beginning. Per/. Cad.

4.

4. But this unalterable harmonic form is, nevertheless, subject
to a rhythmic modification, which consists in projecting the final
tonic chord a beat or more beyond its proper cadence-accent, by
holding over the preceding dominant harmony (or parts of it) in
any of the manifold Suspension-forms. Thus :

5.

Adagio,

I

w.

\^^

:t}*

1^

^i-^

^

^s

Par. 5.

THE PERFECT CADENCE.

r

r^-

^.

t

Beethoven

IT K

:iiiMiJL

J-± I

f ^

n

i

V (V) ♦!)

*i) Thie V at the accent of this cadence-measure is merely a Suspension of
the Dominant which appears in the comparatively unaccented third measure.
This unaccented impression of the Dominant must always be preserved, in order
to sustain the comparative rhythmic superiority (accent) of the cadence- tonic.

The object of the rhythmic modification is, to avoid abruptness, or to
diminish the power of the cadence. As a rule, the cadence-tonic may be
thus deferred to any extent (even beyond the cadence-measure); but it is
generally objectionable for it to appear upon any pulse beyond the final accent
(in case it is compound time, with more than one accent to a measure), or
beyond the last full beat in 2/4, 3/4, and other simple measures. In the
Polonaise, and certain other styles of composition, this belated cadence-tonic
is characteristic. For example :

Allegro moderato

Chopin.

V I V I

See also Ex. 63, — Cadence on the second 8th-note, in 2/4 time.

5. Viewed in its Harmonic aspect, the Phrase will contain
a certain number of different chords between the initial Tonic and
the perfect Cadence, but their number and choice cannot be re-
duced to any rule. In general, it appears wise to use as few
CHORDS AS POSSIBLE (better only one chord for an entire measure,

lO

HARMONIC ASPECT OF PHRASE.

Par. 6»

than a different one for each beat), because the purest and strongest
melodies arise from the simplest and quietest harmonic source.
Still, it is just as necessary to observe the principle of variety with
regard to chord-durations in a Phrase, as with regard to the tone-
durations in the rhythmic structure of a melody. The following
example illustrates the two extremes of harmonic repose and har-
monic activity in the chord-design of a Phrase :

(a) One chord only, from beginning up to Cadence.

i

7.

(

h

-H — f

r

^ J. S ■!:

-T m -H ! — #

«

^=S^

m

5s33:

a

t=t

(I) g-b-d.

Mendelssohn.

t^^r

i

{^

> r- J

f

I

(b) Different chord to each beat (rapid tempo).

Schumann.

I

^f ^>^

Imperf. Cad.

m-a. \ [^ '^w^mrr: r rJ i ' t ^t

The pupil should scan a large number of the Phrases given in this book
(and such others as he may encounter in examining classical music literature),
and closely observe the harmonic structure, i. e., the number and choice of
chords upon which the Phrase is based. See Exs. 6i, 62, 64, 73, 74, 76, 83, 89,
91, 94, 96, and others.

6. The part of the Phrase where the harmonic arrangement
appears to assume a certain degree of regularity (neither by accident

Far. 6.

APPROACH TO THE CADENCE.

II

nor tradition, but for perfectly natural reasons), is the approach to
the cadence.

The final cadence-basstone is the Tonic note, and this is pre-
ceded (in Bass) by the Dominant note of the scale, as has been
«een. The former is accented ; but the rhythmic location and
•extent of the penultimate basstone (Dominant) is optional. It
usually appears twice, at least ; the first time often as J chord of
the tonic (I,).

The basstone most naturally chosen to precede this penulti-
mate Dominant, is its lower neighbor^ not only because of its tonal
importance as Subdominant of the scale, but because of its con-
venient proximity,— conjunct melodic progression being generally
preferable to disjunct. The location, extent, and harmonic char-
acter of this antepenultimate basstone is also optional ; and it may,
furthermore, be either thejegitimate 4th scale-step, or may be raised
'-*«>. Altered tone). For illustration :

Any rhythm.

m

m.

±

Bass part ;

Time-valaes

optiimal.

ppj-^j^tJSUJ ^^

most com-
mon form.

or:

or:

9 : ^ ' "'>

13C

22:

I

-^-

Sabd. Dom — Ton. IIi I» V I IV II

Maestoso

I.

Wagner.

Snbd. Dom. Ton.

12

APPROACH TO THE CADENCE.

Par. 8.

Sometimes the upper neighbor of the Dominant is chosen to
precede the latter, as antepenultimate basstone ; either as legitimate
6th scale-step, or lowered.

Or, both neighbors precede the Dominant basstone, in either
order. Thus (Bass part alone) :

(C major. ^

9.:9^

zz:

I

le:

-o»-

31

E

-^-

-t^-

-^-

VI

IV

if.

7, This is the harmonic ** highway" to the perfect cadence^
as marked by the lowermost part ; and it will be traversed in such
modes, such varieties of style, and in such degrees of speed and
forms of rhythm, as the fancy of the composer (who is not limited
to copying or imitating what others have done, if he will simply
exercise the faculty of original combination) may discover. In the
following instance (an 8-measure Phrase, with an unusual ' ,^
outset — li instead of Tonic), the three basstones shown i
8 underlie the entire Phrase, from beginning to end :

Allegro.

lO.

J.

^^^

»H

f

J I

^^*

^ w * 4 \ f.

Subdom.

(rained) .
Beethoven.

Si/

Dom.

Tonic.

8. Viewed in its Melodic aspect, the Phrase (or, more exactly,
the melody of the Phrase) will be found to consist of a certain
number of sections, called Phrase-members or melodic Members, or
Motives, or (if very brief) Figures. They are more or less dis-
tinctly separated from each other by slight interruptions— corre-
sponding to the ** Cadences" which separate entire Phrases from.

Par. 9.

MELODIC ASPECT OF PHRASE.

13

each other, but differing from these *^ Cadences" in being so tran-
sient as to subdivide the melodic line only, without severing the
harmonic continuity of the Phrase. (Some writers call these spaces
between the members ** Quarter-cadences".)

This element of ** Phrase-syntax " is obviously the most important one in
homophonic phrase-construction, for the entire purpose and signification of
the Phrase is concentrated in its principal melodic line, i. e., its Melody (the
"Air or "Tune", as it is more popularly called). The Melody is the
Phrase, and all the other structural and technical factors subserve the melody.

But the author would reiterate his belief that melody, to a very large
extent (if not altogether), is primarily a Product. The musical Source is
THE Harmony, i. e., the Chords; and the melody is the Product out
OF THIS Source, just as leaf, flower and fruit are the product out of soil,
root and branch. Hence, the student will do wisely, at least as beginner, to
bear in mind the rules of rational chord-succession, and evolve his melodic
motives outof this consciousness. The pages of our strongest classical literature
abound in proofs of direct eduction of melody out of the chord. Examine the
harmonic origin of the melodic motives in Ex. 2, No. 2 ; Ex. 7, a ; Ex. 34, No. 2 ;
Ex. 35, Nv I ; Ex. 46, No. 3; Ex. 74; and others.

At the same time, inasmuch as the melodic extract from one and the same
chord or chord-series may assume a multitude of different shapes (precisely as
one and the same melody may generally be constrained into agreement with
different accompanying harmonies), it is plain that the choice of shape will be
lictated by other conditions altogether, namely : those of thematic relation,
agreement, or contrast, and other structural considerations. Therefore, it is
not only possible, but even necessary, to design, a Phrase largely from the
standpoint of its Melody alone, without running the risk of any serious
violations of the harmonic laws which must, by this. time, be almost a second
nature to the student. Review paragraphs 5, 5, and 7, the principles of -which
must never he neglected or ignored,

9, In rare cases, there is no perceptible break in the Phrase-
melody ; in other words, the Phrase consists of one single member^
in which, at most, a vague subdivision into Figures may be traced.
I. Allegretto. Mendelssohn.

11.^^^

-¥r^-f^^^^

1

4 measures.

a. Allegro

Beethoven.

Ji • I • • • • •

m

m

Usually, however, there is at least one melodic interruption, as
a rule exactly in the middle of the Phrase, dividing the latter into

H

MELODIC MEMBERS.

Par. 9.

two equal members (the pupil may first glance over par. lo). For
illustration :

Allegro.

Haydn.

I.

2 measnres. | 2 measures.

Also Ex. 6i, meas. 1-4; Ex. 70, meas. 1-4.

Additional interruptions are most likely to occur exactly in the
center of either the first half, or (rarely) second half of the Phrase ;
or in the center of both halves. Thus :

X. Allegro. Mendelssohn.

1 measure. | 1 meas. | 2 meas.

First half of Phrase subdivided ; also Ex. 47, meas. 1-4 ; Ex. 72, meas. 1-4. —

a. Moderato. Mendelssohn.

3 meas. | 1 meas. | 1 meas.

Second half of Phrase subdivided; also Ex. 15; Ex. 29. —

Haydn.

Imeas.

1 meas.

I Imeas.

I Imeas.

Both halves of Phrase subdivided; also Ex. 60, meas. 1-4; Ex. 66, meas.
1-4 ; Ex. 74.—

In this process of division (or, more properly, the synthetic
process of compounding figures and members into a complete Phrase)
the principle of regularity naturally prevails ; but there are occa-
sional cases of irregular metrical association, as follows :

I. Andante. Mendelssohn.

14.F

2 beats. | 6 beats
2. Allegretto.

I 2 bents. I 6 beats.

Mendelssohn.

:?=P

-H-f-

p-r

It^^^

Vr^ l JJr l l

4 beats.

I 3 beats. | 4 beats.

I 5 beats.

Par. 10. MELODIC MEMBERS. I5

10. The means of marking (or of locating) the ** spaces'*
between the members of a Phrase-melody, are numerous, but often
so vague that the student will frequently have to depend more upon
instinct than upon rule ; different analyses of the same melody are
often possible, and, in fact, the performer may (to a certain degree)
place an arbitrary construction upon the syntax of the Phrase.

This is simply because the distinctions between the inferior grades of
"Cadence" (in the sense that every interruption, however slight, is a
** Cadence '*), like those of punctuation, are necessarily subtile; and, there-
fore, it is not always possible to define with accuracy the various degrees of
cadential interruption corresponding respectively to the space between Figures,
between Members, or even between Phrases themselves. See par.ii.

The most reliable methods of marking the limits of a melodic
m>em-her^ are :

Istly^ to introduce a Rest (sufficiently emphatic not to be con-
founded with mere staccato\\ this is seen in Example 12, Example
13, No. 3, and Example 14, No. i.

2ndly^ to dwell upon the final tone, thus giving it compara-
tively longer duration than its associates, as in Example 13, No. i, .
and Example 14, No. 2. That this rule cannot hold good for every
heavy note, is demonstrated by the J at the beginning of measure 4
in Ex. 12, and by the first note in Ex. 3 — neither of which marks
the end of a melodic member. The heavy tone must stand in the
proper place, and must impart a cadential (closing, concluding)
impression. Ex. 13, No. i ; Ex. 14, No. 2.

Srdly^ the trait which is more convincingly indicative of melodic
division than any other, is that of Repetition or Sequence^ or,
in a word, the recurrence of any sufficiently striking melodic or
rhythmic figure. See Ex. 13, No. i, members i and 2 ; Ex. 13, No.
2, members 2 and 3; Ex. 13, No. 3, members i and 3, 2 and 4;
Ex. 14, No. 2, members i and 3; Ex. 17; Ex. 18; and the follow-
ing, in which the characteristic repetition of the first tone unmis-
takably marks the beginning of each following member :

IB.:

x-^=::=^_^

The only danger of error, in this method of analysis, is that of
accepting too many •* spaces", and thus confounding the smaller
particles of the Phrase (the Figures, or subdivisions of the Mem-
bers) with the Members themselves. Further :

i6

MELODIC MEMBERS.

Pw.li

11. The smallest musical particle is the single Tons,-
responding to the single letter in orthography. The uninterrupted
association of tmro or more tones (in melodic succession) constitutes
the Figure.

The association of Vko or more figures (separated by very
slight interruptions) generally constitutes the Phrase-member or
Motive, — corresponding to the words, small and large, in a sen-
tence.

The association of two or more members (separated, as above
shown, by " quarter-cadences ", corresponding approximately to
the spaces between the w^ords of a printed sentence) generally con-
stitutes the complete Phrase or sentence.

The usual (but not invariable) subdivision of the melodic mem-
ber into its figures, and the consequent distinction betiveen these
factors, is shown by the double system of slurs in the following
illustrations (upper slurs for Members, lower ones for Figures) :

1. Presto.

Mendelssohn.

16. r

a. {Ex. 12,)

♦i) These possible variations in the analyses of Phrase-syntax need give
the student no concern. Review the first few lines of paragraphs 8 and lo;
and examine the given examples with regard onlj to such of their divisions
and subdivisions as are plainly definable.

1 2. In the composition of a Phrase-melody, some attention
must be paid to the melodic and metric relations between the
members ; and while there will appear to be no system, nor any
limit to the possibilities, still, each individual possibility is a
suggestion to the alert student, and consequently becomes a definite
stimulus to his imagination.

As concerns the metric relation, it has already been seen, in
Examples 12, 13, and 14, that the Phrase-members may be either
similar or dissimilar in length.

\

Par. 12.

MBLODIC MBMBERS.

17

And with regard to their melodic (or thematic) comparison,
both similarity (Ex. 13, No. 3) and difference (Ex. 14, No. 2) have
been observed.

In Examples 12 and 13, similarity of rhythmic character pre-
vails, while in Example 14 there is striking rhythmic diversity.

There is probably a slight preference in favor of regularity and
similarity, in all of these respects.

Total thematic agreement (Repetition) is illustrated in the
following :

X. Andante. MENDELSSOHN.

I

I

17.i^^^

«-

^:

*-*^

liJ.-

I

fr^j J f =^

2. Vivace,

a.

Haydn.

jfe fe f i ' i rr^

^^

See also Ex. 30, No. i, meas. 1-2; Ex. 35, No. i, meas. 1-3; Ex. 46, No. 3;
Ex. 71, meas. 1-4.

Comparative thematic agreement (by Sequence) prevails in the
following; (N. B. A Sequence is the reproduction of a melodic
figure or member upon other ^ higher or lower ^ steps) :

I. Vivace.

^ \^ ^1

i— H k»-T= i-r= h

18.

Haydn.

: y.vr i rr_jJ'c/K^r»:tnLi'rrf;i^^ ^

a. Presto.

Mendelssohn.

See also Ex. 37, meas. 1-3; Ex. 34, No. i, meas. 1-3; Ex. 37, meas. 1-3;
Ex. 52, meas. 3-5.

Frequently, actual thematic agreement is cleverly disguised, or

modified, as follows :

X. Presto. Haydn.

- I « .^ « primary form. " '^^

i

\J ¥

a. Seq. mod.

/

J'

/

z8

9. Andante*

MELODIC MEMBERS.

Mendelssohn.

Par. 12.

« prim. form.

3. AlUgro,

m

^^

^ I I a. embelliflhed.

Brahms.

4. Allegro.

Chopin.

t

f^

Haydn.

i

a.

contracted and shifted.
See also Ex. 30, No. i, meas. 2-3; Ex. 79, meas. 1-2, and 5-6.

After carefully and repeatedly reviewing all the conditions of
Phrase-construction given in the preceding paragraphs, the pupil
may venture to apply them in the invention of original examp^les of
Phrase-melody, according to the following directions :

EXERCISB /.

Invent a number of instrumental Phrase-melodies of diversified character
and design, principally four measures in length, — a very few of two and
eight measures.

Alternate regularly between the major and minor modes in successive
examples ; employ the different varieties of duple and triple time impartially ;
and employ the various grades of tempo (from Adagio to Allegro). Exem-
plify the principal forms of phrase-syntax (Exs. 11, 12, 13), and thematic
relation (Exs. 17, 18, 19).

The harmonic basis must be borne in mind constantly (par. 5, 6, 7), but
is not to be written down ; the Melody alone constitutes the object of this
first exercise.

Par. 14a.

HARMONIC ECyUIPMENT OF THE PHRASE.

^9

CHAPTER II.

THE HARMONIC EQUIPMENT OF THE PHRASE-MELODY.

1 3. The general principles governing the harmonization of a
given melody having been acquired in the course of *' harmonic '*
study, it only remains to add a few directions here in reference
to the various styles of harmonic accompaniment. These will
be dictated, most naturally, by the character and tempo of the
melody ; but they may be chosen according to the particular effect
desired.

1 4a.. In those cases where the melody in itself exhibits its
harmonic source with sufficient distinctness, or where it is to appear
in impressive isolation, the harmonic accompaniment is apt to be
omitted altogether, during one or more members, or perhaps during
the entire Phrase ; the melody appears as solitary part, or tone-
line (unison), or is simply doubled in one or more octaves. For
example :

I

I. Allegro moderator

Schubert.

I

J-. ' J-.^U.

20. {

fV

m

^■ 4 <g ^

■^ ^IJ.!!* ^ 'rt^^'H ^^

rri y

r

a. Allegretto,

iJ*V J

etc.

r^"Hf

/

20

HARMONIC EQUIPMENT OF THE PHRASE.

Par. 14b.

Chopin.

3. Lento.

(one part

^^

See also Ex. 38, No. 6 (also 4 and 5) ; Ex. 52; Ex. 81, meas. 1-2; Ex. 84,
meas. i ; Ex. 94.

(b) Or, the Phrase-melody is supported by one additional
PART or tone-line, which either assumes a coordinate melodic
character (Nos. i and 2), or becomes, as figural part, the arpeggi-
ated representative of the entire chord-basis (Nos. 3 and 4). Thus :

I. Allegro,

Mendelssohn.

W- Jl^:^7f

21. \

J:

m

rmii

i^m

*-*

± S *:

f

(2 parts, doubled in 8ve8 )

3. Vivace,

Haydn.

(^

fe

^^

?=^

LiiHiJji

T jjj i J ''''{jM ^^ n

(2 equal parts.)

3. Allegretto.

Beethoven.

Par. 14c.

HARMONIC EQUIPMENT OF THE PHRASE.

21

4. Allegretto.

mr^—^

Schumann.

E

(flgnral part. )
See also Ex. 32, No. 3; Ex. 64; 68; 69; 77; 78; 97.

(c) The addition of two accompanying tone-lines (3- voiced
style) secures a complete harmonic effect without bulkiness, and,
for this reason, the 3- voiced style must be regarded as generally pref^
erahle and most commendable.

Here, again, the parts may be similarly melodious, as in the
vocal style (though dictated by the rules of c^or^-structure) ;

Or one of the tone-lines may be a figural part, — a choice
which is more common and appropriate in instrumental (or key-
board) music, and especially suitable for the Homophonic forms.

In this harmonic style the device of duplication in thirds
or sixths is both convenient and effective (see particularly Ex..
22, Nos. 4 and 5 ; and observe how the parts are related in
Ex. 62).

For illustration :

1. Adagio.

Mozart.

H\ r ■• Li' J-j

# t r ^ " I

IS^

S |i I I I f

^

rt

etc

Mendelssohn.

T| / 8 pg-^

•^^ m . '^^rn

etc.

^

i

22

HARMONIC KQJJIPMKNT OF THE PHRASE.

Par. 14d.

3. Andante con moto.

Mendelssohn.

I ^ii ^ jg ^ jg ^ J^ ^ /J 'i

marc.

J

/

etc.

f

tt:u

I

4. Andante,

Mozart.

|j^^^

:p=^

j^

irirJ ;i»^-fe

^

f ^ M

(Duplication of principal mel. in thirds.
5. Allegro.

nr r ,_ r- 11

Beethoven.

See also Ex. 26; 28; 49; 56; 58; 59; 66; 75; 89.

(d) The 4- VOICED STYLE is often necessary for greater harmonic
breadth and fullness. It usually resembles the vocal (or choral)
style very closely (Ex. 27), although one or more of the four
tone-lines may be a figural part. It will be most appropriate for
melodies of a comparatively serious and stately character, but less
desirable for graceful or rapid melodies. Illustrations of this style
will be found in Examples *ja and 3, and in the second half of
Example 5 ; to which may be added the following varied specimens
of its treatment : ft* I I

1. Andante. \ \ \ 1*^ ± I2J. JL ^^^'^"-

m

vmj

^

I I I

§

I I I I I f f — f -

etc.

I

(Imitation of vocal style.)

PM. U. harmonic EqiJIFMSNT OF THS

. AiUgra. Bbeth. j. Andaittt,

^ •»?

Se«a1soEz. 35; Ex. 2f\ 36; 51, No. 1; 54; 72; 74; 96-

1 5. By simply duplicating one or more of these fundamental
tone-lines (or single chord-intervals) in upper or lower octave-
registers, the harmonic volume may be increased to any desired
extent, and copious, rich, powerful or pompous effects achieved.
Such voluminous harmonies are seldom sustained very long ; and it
is very important to observe that, in general, (only excepting in
strict vocal writing,) it is neither necessary nor -wise to adhere
rigidly to any number of tone-lines. That is to say, the volume of
harmony may be increased or decreased during certain members or
figures, or even at single points, for the sake of dynamic and har-
monic variety. But the principle of "coherent tone-lines" {the
most vital in music) must never be violated; at least two lines (the
uppermost, or Melody proper, and lowermost, or Bass) must always

24

HARMONIC EQJJIPMENT OF THE PHRASE.

Par. 15.

be conducted with strict regard to melodic law ; and usually one
inner part also. These quantitative fluctuations in harmonic volume
should, therefore, plainly appear to be transient duplications, or
omissions, of a portion of the original bulk, which do not disturb
the effect of the fundamental lines (** Material of Mus. Comp.'\
par. 413, 414).

For illustrations of this species of harmonic accompaniment,
and also those explained in par. 14, the pupil is referred to the
pages of standard literature, (Mozart, Beethoven, Mendelssohn,
Chopin, Schumann, Brahms), which he is urged to examine with
special reference to this point. See Ex. 2, Nos. 2 and 3; Ex. 5;
Ex. 6 ; Ex. 10 ; Ex. 20, No. 3 ; and the following :

I. Maestoso. Mendelss. a. Agitato. Brahms.

i

«

2^

mf %

mi v ^ ; ii i j' p i

etc.

etc.

A

r
I

3. Allegretto.

^

Beeth.

*• ■» <■ ■»• -te .

^

V

etc.

^^

V ■> ■/ >

See also Ex. 38, No. 3; Ex. 40, No. 4; Ex. 47; 61 ; 67; 83; 91 ; 95.

MISCSLLANEOUS EXAMPLES OF THE SIMPLE 4-MEASURE PHRASE.

Schubert, Songs: ** Die schOne Mlillerin ", first four measures of Nos. r,
4, 6, 8, 10, 14, 20; ** Winterreise", first four measures of Nos. 9, 11, 14, 22, 23.

Mendelssohn, ** Songs without Words", No. 4, meas. 5-9; No. 28, meas.
1-4; No. 41, meas. 1-4; No. 44, meas. 1-4.

Also Ex.62, meas. 1-4; Ex. 66, meas. 5-8; Ex. 73, meas. 5-8; Ex. 74, meas.
1-4; Ex. 76, meas. 1-4; Ex. 78, last four measures; Ex. 92, meas. 1*4; Ex. 96*
meas. 1-4.

Par. 16. THE DEVELOPMENT OR EXTENSION OF THE PHRASE. 25

EXERCISE 2.

Add the harmonic accompaniment to some, or all, of the Phrase-melodies
invented as Exercise i. As a very general rule, the pupil is expected to adopt
the f ianof or testy le^ at least for a time ; though he is at liberty to write
occasionally for any other instrument or ensemble with which he may be
familiar (Organ, String-quartet, Trio, Duo, etc.). He is warned not to over-
step the purpose of the Exercise, which in this case is limited to the simple
regular Phrase.

Though it is by far the most common habit to place the Phrase-melody in
the uppermost part, the pupil is urged to assign it now and then to an inner
part, or to the lowermost one. See Ex. 22, No. 3; Ex. 28; Ex. 91, last seven
measures.

Review paragraphs 5, 6 and 7, and avoid overloading the phrase. The
FEWER CHORDS, THE BETTER. Avoid the faulty habit of harmonizing eack
individual melody-tone with a separate chord ; that is, harmonize as many
successive melody-tones as convenient with the same chord ; taking care, how-
ever, to change the harmony at the stronger accents, i. e., the first accent in a
compound measure, and, analogously, the first accent of all measures which, in
the regular alternations of heavy and light pulses throughout the Phrase, rep-
resent the heavier units.

Very positive preference must be given to the 3-voiced style.

CHAPTER III.

THE DEVELOPMENT OR EXTENSION OF THE PHRASE.

1 6. Musical composition involves two distinct mental pro-
cesses : that of Conception and that of Manipulation.

The process of conception comes first, and dominates during
the creation of the thematic germ, the Motive or Phrase. The
process of manipulation is maintained more or less steadily the rest
of the way, even though it be in such close interaction with the
imaginative and conceptive faculties that the limits cannot always
be defined in the writer's consciousness ; upon this latter process
depends most largely the success of the product. Composition is
not an aggregation (merely the collecting and associating of a
quantity of kindred, or, worse still, heterogeneous musical frag-
ments, without regard to structural design, logical arrangement, or
unity) ; it is the result of evolution and logical deduction, — ^an

26 PHRASE-REPETITION. Par. la

unfolding of one phase after another out of the thematic germ,
urtil the growth is consummated.

Of this fact, the pupil should encourage the most absolute conviction,
for any other conception will prove an effectual obstacle to his aspirations in
the classical avenues of serious and enduring musical composition.

1 T. The ** Phrase " having been conceived, the next step, then,
is its enlargement or development. This is not yet to be effectu-
ated by the addition of other Phrases, for that would overstep the
limits of the present purpose, which is : the development of the
resources embraced within a single germinal Phrase, upon a cor-
respondingly narrow scale of structural design.

The principal means employed in the Development or Exten-
sion of a Phrase (or thematic germ of any kind) are those of
Repetition, Sequence (i. e., the reproduction of a figure or
member bodily, a certain interval-distance higher or lower), and
Expansion. Their application may be classified in a fourfold
manner, namely :

(i) The Repetition of the entire Phrase;

(2) The Extension at the End,

(3) The Extension at the Beginning, and

(4) The Extension in the Course of the Phrase.

I. Phrase-repetition.

1 8- When the whole Phrase is to be repeated, it is customary
(though not absolutely necessary) to fill out the measure allotted to
the perfect Cadence, by continuing the rhythmic pulse in some form
or other, so as to ** bridge over" the space between the end of the
Phrase and the beginning of its repetition, and thus partly conceal
the Cadence. This is more or less desirable, according to the width
of the intervening space.

It is not to be an '''' Evasion* ^ of the perfect Cadence, and there-
fore the harmonic form and the cadential effect of the latter must
be preserved.

The final Tonic chord must be retained; but it may be
embellished ;

And the chord-third, or even the chord-fifth, may be substituted
for the Root in Soprano, or (rarely) in Bass.

The rhythmic pulse should be maintained within th^ tones of
the Tonic harmony ; still, smooth harmonic progressions (through

Par. 19.

PHRASE-REPETITION.

27

other chords) may be ventured, if made in so cautious a manner
as not to cancel the cadential impression (i. e., not until the toiitc
effect is established). See par. 21 and 25a. For example :

z. Andante.

*i) Correct perfect Cadence, omitted because of the coming repetition.

*2) Concealed form of perfect Cadence : compare with final measure. The
rhythmic pulse is actually more rapid during this measure (i6th-notes) than
anywhere else in the Phrase ; and the uppermost part has the chord-fifth (e)
for a time, instead of the root (a), as at the end. The cadential impression is,
however, preserved by the cessation (rests) in the lower parts. For other
illustrations of cadence-bridging, see Exs. 26, 27, 28; Ex. 46, No. i, meas. 4;
Ex. 50, meas. 8; Ex. 66, meas. 4; Ex. 73, meas. 8; Ex. 78, meas. 12; Ex.
89, meas. 8; Ex. 92, No. 2, meas. 8.

] 9. The Phrase may be repeated literally^ without any
changes whatever, excepting those involved by concealing the per-
fect Cadence. See Ex. 25.

Also Beethoven, Pfte. Sonata, op. 27, No. i, first movement, first, second,
and third Phrases (repetition-marks); Pfte. Sonata, op. 28, Finale, meas. 1-8;
9-16 (slightly embellished). — Schumann, "Album-Blatter '*, op. 124, No. 2;
No. 7; No. 9; No. 15; No, \*j— first eight measures of eacA.— Chopin, Prel.
op. 28, No. 15, meas. 1-8.

But it is far more usual and desirable, as conducive to thematic
development, to introduce unessential alterations^ or variations,
during the repetition. (See par. 40.) By *' unessential" is meant
unimportant — not affecting any characteristic or essential element

28

PHRASE-RBPETITION.

Par. 19b.

of the Phrase. These unessential modifications of the repetition
may be classified as follows :

(a) Unessential embellishment of the Melody :

Andante.

26. <

I

9? ^

•*— ^

,^ n^.-j^j^. ^^

^

* r~?^ •

^^p

ttzfi:

I

trf-T7^

t:

-^M-

c«=?

i

I -,

(modified repetition.)

. ! . . . i I

a^

9 ^.U-l^

^

Beethoven.

*i) In this concealed form of the perfect Cadence, the cadential impres-
sion is preserved by every prescribed condition (par. 3) excepting the pause or
check in the melodic current; the rhythmic movement is here again (as in Ex.
25) accelerated, and thus maintained during the melodically ornamented repe-
tition. See also : Chopin, Mazurka No. 2 (op. 6, No. 2), first 8 measures.

(b) Changes in the harmony and modulation :

27- <

Par. IDc.

PHRASE-REPETITION .

29

I

<P

tition.)

-1-

Folk-song.

i

SS

i

f

>

1

^ ^ n

r U ' r ' f "t

gs

4 4 iiu i J i. J _ ,n J

t=t

^1^

:j i r f

i

r

See also Ex. 51, No. i, meas. 1-8. — Ex. 62, meas. 9-18.

To this class of modifications would belong, also, a repetition

in the opposite mode of the same keynote (i. e., transformed from

major to minor, or vice versa) ; returning, perhaps, to the original

mode at the final Cadence.

See Chopin, Nocturne 17, meas. 14-6 from the end; Chopin, Mazurka
20, meas. 17-24; also last nine measures; Beethoven, String-quartet, op. 18,
No. 6, first movement, meas. 45-52. Ex. 89, Part II, meas. 1-8.

(c) Shifting the Phrase-melody upward or downward to a differ-
ent register, or different part; either bodily (i. e., the entire
Phrase) , or in sections :

AlUgro.

B#^

» I V 1 * - ^ f P

Ff^

^

# P I — ^

(concealed
cadence.)

W^^

s

fc=:5:

E^^

1^^ I

y-t=

e

^y-FTH

5

■«-^-

30

PHRASE-REPETITION.

Par. 19d.

Mendelssohn.

*i) The melody is shifted bodily upward one and two octaves, from
"Tenor" into "Alto and Soprano" registers. See also Exs. 29; 30, No. i;
31. Beethoven, Bagatelle, op. 33, No. 6, meas. 1-8.

(d) Changes in character and style of accompaniment :

Allegro,

29.

i

Repetition'.
1( Violin) J

J u J J

^£&a ^6b ^^ S^ ^ &^

I

t,

^

i

? — r r

Beethoven.

^

J a ^

^S

J.

i

i

§

^^^

?^^b:ts ~S

"^

r

3

^

^5

3

*

s

8

*i) Here the Cadence is not concealed, because the " space " is so narrow
as to render it unnecessary. See also Ex. 30, No. i. Bbbthovbk, Pfte. Sonata,
op. 54, first movement, meas. ,70-77, and 106-113.

Par. lOe.

PHRASK-RKPETITION.

31

(e) More complete and radical changes in the course of the
Phrase-melody (not, however, affecting the beginning, or the end,
or destroying general resemblance) :

z. Andante.

£ • fgff :

30.

^^

'D

^^^^^^^^^^^^^

^^^-'-

(repetition.)

■^ ^

Beethoven.

9 • W ir

-wh-^—^-^f—^

a. Allegretto.

I

te

tt

S

^-^

I2p:

■^-

^m.

^FJF^

-> ^ * * J J

yr

(repetition.)

Haydn.

1^

t^

:e

rffrrrir

^

i

^

*2)

I

t

r ?' /i^' r r ^^y

I

Ft f

*i) This Cadence is not concealed, before the repetition, because there is
no space left at all. See par. 4.

*2) The second member of the Phrase is entirely changed ; but the Cadence
is preserved. See also Ex. 51, No. i, last eight measures.

T

32

PHRASE-REPKTITION.

Par. 30

20. In rare cases, the bridging-over of the perfect Cadence,
before repetition, is extended into a brief Interlude, or transitional
passage, of one or more (superfluous) measures; but it must be
kept so inferior in character and contents as to appear unessential
and extraneous. For illustration, the 8-measure Phrase given in
Example 10, is continued and repeated as follows :

31.

(measure 7 )
(Ex. 10.)

(Cad. measure.)

(nnperfliioiis nieasare.)
*1) (luterliule.)

^irta

9^

4-

^1

(Repetition.) *2)

Beethoven.

*i) This measure is kept subordinate hy reduction to one part, and by the
suppression of the accompaniment.

*2) Compare this repetition with Example lo; it illustrates the change
of register (19c), not entire, but in sections.

See also Mendelssohn, ** Songs without Words", No. 46, meas. 3-12.

Nothing more than such a distinctly unessential interlude could
intervene between a Phrase and its * ' repetition ' ' , without trans-
forming the structural trait of "repetition" into that of *' recur-
rence ". A repetition is a (practically) immediate reproduction.
See par. 81 a and b.

Par. 23. PHRASE-REPETITION. 33

2 1 . Reverting to par. i8 (which review), it must be dis-
tinctly understood that no essential change of the harmonic form of
the Cadence is permissible in the case of entire repetitions. This
preservation of the cadence-harmony is the inviolable condition
upon which the assumption of * ' repetition ' ' depends.

The Cadence is the ** aim'* of the Phrase ; and it is only when the original
harmonic aim is retained, that the Phrase niay be said to have been "repeated".
Any essential alteration of the harmonic aim (the Cadence-chords) would
resolve an otherwise apparent "repetition" into two separate Phrases, with
independent direction and aim. See notes and context to Ex. 46. The dis-
tinction between apparent repetition and actual repetition (depending upon
whether the Cadences are essentially different ox not) is of the utmost impor-
tance, because genuine repetition, whether variated or not, never constitutes
any change or actual advance in the formal design of the composition;
whereas, an apparent repetition, which proves ultimately to have been
thwarted by a real change of cadence, contributes, as new member, to the
progressive enlargement or evolution of the design.

22. Two, or even more, repetitions may occur in succession,
somewhat like a series of simple Variations (e. g. , the Ciaccona for
solo violin, by Bach ; the Chaconnes of HAndel, for Harpsichord ;
the thirty-two C-minor Variations of Beethoven, etc.) ; each repe-
tition being differently modified, possibly in a certain systematic pro-
gression, See Beethoven, Bagatelles, op. 33, No. 6, meas. 31-46.
Brahms, op. 118, No. 5, second tempo (Phrase with 5 repetitions).

23. The OBJECT of Phrase-repetition, and of repetition in
general, is :

To establish unity of thematic design by direct corroboration ;

To give additional emphasis to the Phrase or member, and to
defjne its contents more clearly ;

And, finally, to obtain greater simplicity and repose of elemen-
tary character than would result from a succession of constantly
changing Phrases or members. The repetition will, therefore,
be most desirable when the Phrase is of a somewhat abstruse
or complex nature ; or of unusual length ; or when interesting
and important enough to tempt the desire for (and justify) a second

hearing.

EXERCISE 3.

Select a number of the Phrases invented in Exercise 2, or invent new
ones, and add a repetition to each, illustrating the five principal varieties of
modification given in paragraph 19. All five modes may be applied successively
to one and the same Phrase (22) ; or single repetition may be applied to e^ch of

34

THE EXTENSIONS AT THE END OF A PHRASE.

Par. 25a.

five different Phrases, with that variety of modification which appears to be
respectively most appropriate.

2. The Extensions at the End of a Phrase.

24*« The extensions at the end are of a two-fold nature, con-
sisting, istly, of such as lie within the Cadence, and 2ndly, of such
as lie beyond the Cadence.

. To the first class (inside) belong :

(a) The repetition of the second half of the Phrase ; possibly
exact, but usually modified.

This involves, as a rule, a more or less complete ** Evasion "
oy the perfect Cadence^ — an act which differs from the simple con-
cealment of the Cadence (shown in par. i8), in being a more pro-
nounced deviation from the latter, and often consisting in a different
form of harmony altogether. The approach to the Cadence is usually
not affected by the prospective evasion, as the latter does not begin
until the Tonic element of the Cadence is due, or has been felt,
upon its proper beat. The evasion is then effected :

By substituting an Inversion for the fundamental form of the
Tonic chord (i. e., 3rd or 5th in Bass) ;

By substituting the VI for /the I (Ex. 33, No. i) ;

Or by substituting any o^er chord, even of another key, which
contains the original Tonic note in sufficient prominence to preserve
the cadential impression.

In any case, the rhythmic pulse is sustained in one or more
of the parts ; and the melody and harmony are so conducted as to
lead smoothly into the desired repetition.

The difference between "concealing" and "evading** the Cadence
will become amply apparent upon carefully comparing the details enumerated
in these last few lines, with those given in paragraph 18.

For illustration :

z. Andante.

9^

^r#=t

I

evaded

te

B

Par. aSa. THE BXTBXSIONS AT THE BXD OP A PRRASB.

35

Mendelssohn.

3. Andante.

W=^

fc=d

k^

H-

^^

A=t

i

¥

i^

^ ^ s^ ^

evad.

^^^M

Mozart.

f

3l3t

'

ft

Cad. »4)

^^^

^^

# I ^

^

36

THE EXTENSIONS AT THE END OF A PHRASE.

Par. 85b.

*i) The expected Cadence is evaded in this case by placing the chord-
third in the Bass, and the chord-fifth (leading downward to the Root) in
Soprano. Compare with correct form, two measures later; and observe the
modification in the treatment of second Phrase-half, when repeated.

*2) This peculiar evasion of the expected Cadence will be best understood
by comparing it with the correct form, four measures later. It even affects
the approach to the Cadence, but the cadential relation is preserved by cling-
ing to the key-note (e) in the Soprano.

*3) In this example, there are two repetitions of the second half of Phrase;
the first time with the same form of evaded Cadence, and the second time
with the correct perfect Cadence.

*4) Also a peculiar mode of Cadence-evasion, sometimes called ** Inter-
cepted Cadence". It answers sufficiently well for the expected Cadence, as
the accented chord (the IVi — b, d, f) actually contains the Tonic-note. The
repetition covers a little more than half the Phrase, reaching back abruptly to
the beginning of the 2nd measure.

See also Mendelssohn, ** Songs without Words ", No. ii, meas. 9-3 from
the end. Beethoven, Bagatelle, op. 33, No. i, last nine and one-half measures.
Schubert, Song, ** Haiden-Roslein '% meas. 5-10.

(b) The repetition, or (more rarely) the sequence, of the last
member of the Phrase ; possibly a single repetition, but usually
twice in succession.

This generally refers to the last quarter of the Phrase, or a
trifle more ; and the evasion of the Cadence is necessary, because,
in this case, the original (expected) Cadence is not to be included
in the extension. For example :

z. Andante.

•^ (el)

«

i

^S

Cad.

Ext.

Cad.

Mendelssohn. No. 43.*2)

evad. ♦!)

fv Ivi

2. Moderate.

9k, evad.' ♦!) a. evad. a.

^ ^ J i J J I fr] \ K?^ ^' rpi j-^

r

r

I

^y

»4-

^SJlu

I

I

blVI

Par.

THE EXTENSIONS AT THE END OF A PHRASE.

37

Gad.
evad. *3)

Mendelssohn. No. 85. *2)

■Extension.

•i) The expected perfect Cadence is evaded by substituting the VI for the I.
The three notes in Soprano, which immediately follow, are 'included in the
evaded Cadence ^ and serve: istly, to sustain the rhythmic pulse; 2ndly, to
lead smoothly into the first tone (b) of the member to be reproduced. The
reproduction of the last member, which then follows, occurs twiccy each time
in a modified melodic and rhythmic form (the harmony remaining the same),

♦2) The numbers after Mendelssohn's name invariably refer to his ** Songs
without Words".

•3) The perfect Cadence is completely evaded by turning back abruptly one
measure, where, on the first beat, a Tonic chord occurs in the very form suit-
able for evasion (chord-fifth in Soprano, and chord-third in Bass ; see Ex. 32,
No. I, meas. 4). The extension consists in repeating this third measure twice,
the first time exactly, and the second time with altered harmonization.

26. To the second class (outside of, or beyond, the Cadence)
belong :

(a) The Repetition of the entire Cadence-group (of chords),
usually at least twice, and possibly modified.

This will usually embrace the last three or four chords, in-
clusive of the original Cadence ; and no evasion of the latter will
be necessary or even practicable, because the Cadence, as it stands,
is to be included in the extension. The rhythmic pulse may, or
may not, be sustained. For example ::

X. Andante.

sion.)

Beethoven.

E^

±

^3

- \ \j hJ'j UJ \ \ r.uu-\''\\\ \ b'm

^^ ^^^ SSSS5 SSC^

38

THE EXTENSIONS AT THE END OF A PHRASE.

Par.aStx

t

*i) This is a Semicadence (par. 36) instead of a perfect Cadence ; but that
makes no difference, as this principle applies to every variety of Cadence.
The rhythmic pulse is not marked, contrary to common usage. The repetition
of the Cadence-member occurs in this instance no less than eight times, includ-
ing one change of register, an acceleration (diminution) of the rhythm, and a
series of rhythmic shifts which alter the location of the figure in the measure.

*2) The five i6th-notes which follow the Cadence merely serve to sustain
the rhythmic movement (compare Ex. 33, note *i)). — ^*3) Original melodic
form of Cadence, transferred to next higher octave. — *4) Embellished form
of the same. — ^*5) Half-measure sequences of foregoing figure.

See also Ex. 41, No. 4, note ♦7). And Mendelssohn, "Songs without
Words", No. 20, meas. 12-5 from the end.

( b) The repetition of the two Cadence-chords (V and I) ;
usually at least twice, and either in the same form, or with any
rhythmic or melodic modification of the form.

This species of extension will often be found to differ but
slightly from (a), cited above; though it is generally more frag-
mentary than the latter. For example :

\

Par. Mb. THE BXTBNSIONS AT THE END OF A PHRASE.

39

(

*

-) J-. i

Ifel

Mendelssohn.

i

*-*—?-

Cad. ExteDsion.

^i ' '^f^

Tx ?

i

VV IVVIVV

a. Allegretto.

l=i

S

iw

^

J

#t=^

m^

f^-i i 4.qy I s^ ^ ifei?? TTT ^^^^^

g

I

Gad. Extension.
*2)

Mendelssohn. No. 42.

i

^r^¥

W\

r

r

f i_ JL.

r

r

f

^¥->^-^i^^=^^ ^ ^ ^1=^'" ? ^^=^-!^ ^ ^=3=B

^"

^-

I (V I) V

3. AUegro.

(V I)

i

e^^

^

pfe

f

f

f=F

?

£

P

3

*2>-^~jL

^ ^

-# —

Haydn.

40

THE EXTENSIONS AT THE END OF A PHRASE.

P«r.2Bc.

*i) Two exact repetitioiu of the last three (Cadence-) chords.

*2) Four repetitions of the two Cadence-chords, as continuation of the
accompaniment. The first and third repetitions are slighti j altered in form, and
sound parenthetical, because the melodj pauses meanwhile upon its Cadence-
tone. The second and fourth repetitions are an exact reproduction of the
Cadence, including the Soprano.

*3) The Cadence is concealed bj the quicker rhjthm in the Soprano (bor-
rowed from the Bass of the first measure).

See also; Mkkdklssohn, ''Songs without Words", No. i6, first three
measures, and last four measures.

(c) The reiteration of the final Tonic chord, to an optional
extent, and in optional rhythmic and melodic form.

The extent (number) of these reiterations can hardly be de-
' termined by rule, as this extension usually represents the last oscilla-
tions of the swinging harmony, the " dying out" of which will
depend upon the peculiar circumstances embodied in the Phrase.
Instinct will be the best guide. As a rule, the Cadence-tonic, in
this case, must fall upon the Jirs^ accent of the measure. For
example :

36.

Cad.

(Extension.)

♦ V

Schubert.

See also: Mendelssohn, "Songs without Words", No. 4, first five
measures (compare with last five measures); No. 6, last six measures; No. 8,
last eight measures; No. 12, last twelve measures (4-measure Phrase, repeated
and extended); No. 27, last twelve measures (illustration given in Ex. 32,
No. 2, with reiterations of last chord); No. 31, last five measures; No, 42, last
eleven measures (illustration given in Ex. 35, No. 2, with reiterations of final

Par. 27.

THE EXTENSIONS AT THE END OF A PHRASE.

4J

Tonic-chord, and one additional final announcement of the perfect-Cadence
chords, V-I).

(d) A plagal cadence, of more or less elaborate character, dur-
ing the prolongation of the final Tonic note in Soprano, or in both
outer parts. For illustration :

Adagio.

37.

|_£xten8ioii.
See 29 6.

^

m.

fe

^

m

.Cad.

Mendelssohn. No. 44.

<

I

it=g

^P3

Plagal extension.

m.

I

\-

^^

r

1 J

i

-z^

y^N

w

■^^

i

Xl;

f-

7

See Chopin, Nocturne ii (op. 37, No. i), last five measures; during the
plagal extension, the major Cadence is substituted for the original minor.
Nocturne 7 (op. 27, No. i), last eight measures ; a two-measure Phrase, repeated,
and extended by a somewhat elaborate plagal ending.

These various extensions are often applied successively to one
and the same Phrase : see Beethoven, Pfte. Sonata, op. 53, Finale
(Rondo), measures 191 to 220: — (an 8-measure phrase), beginning:

27. The OBJECT of extensions at the end of a Phrase is, to
prevent the movement of harmony and rhythm from breaking off
too abruptly; and the degree of momentum which the sentence
may have acquired, must determine the propriety and extent of their
use. Furthermore, they also serve, of course, to develop their por-
tion of the resources of the Phrase, and are for this reason generally
permissible.

42 THE EXTENSIONS AT THE BEGINNING OF A PHRASE. Par. 28.

JBXBRCISB 4.

A number of former Phrases, or new ones, with extensions at the end,
according to the modes of treatment described in par. 25a and 25^.

EXERCISE 5.

The same, according to the modes of treatment described in par. 26, a, 3,
Cj and d,

3. The Extensions at the Beginning of a Phrase.

28. An extension of the Phrase at its beginning will neces-
sarily assume the nature of an introduction; i. e., some form of
harmony and melody which suggests (or prepares for) the initial
member of the Phrase, and leads smoothly into it.

The chief difficulty is, to keep such an introductory extension so
inferior and unessential in character, that it will not be mistaken
for the real beginning of the Phrase.

The necessary inferiority will be ensured if either the element
of definite melodic or definite harmonic progression be absent, or
obscure, during the introductory passage.

Hence, the best form of introduction is that which is limited to
the figure of the accompaniment, which may anticipate the announce-
ment of the melody proper, either w^ith or without changes of
harmony.

It may, however, also consist in an anticipation of the first
figure of the melody, without accompaniment ;

Or, in a brief one-part *' Cadenza", leading into the latter,
(see Ex. 31, meas. 2 and 3) ;

Or, in one or more simple annunciations of the key-note (or
Tonic harmony), or possibly both Tonic and Dominant tones, or
chords, without definite melodic form.

The length of the introduction is largely optional, but it will

rarely exceed half the length of its Phrase. For example :

I. Andante, Mendelssohn.

Introd II Phrase.

i

(

fe— ^_ , 1 i j. , ^'j ,1 I I

etc.

21^ , p 'Ti , p JT] I , p .7] , pJTlft

r X r j« r

Par. S8. THB EXTENSIONS AT THE BEGINNING OF A PHRASE.

43

3. Adagio.

Introd II Fbnua.

Mendelssohn.

^

it

•^ X y

=^S
?9^

3

i

etc.

"T^* ' I ' -;-f

» > > >

LLTE

i

I

4. Allegro.

»3)

^

3

(_^)-

^ — »-

t=(

^

Introd.

^feS^^^B

_i,; jn

8ib^

K^

• '' ^ h ^ ^-

Gfe f.

Chopin. Op. 51.

Phrase.

ft

?=s=t:

=1=

tzt

t

* ^ f \ T

jDl

etc.

^

■ i\JlU\^

5. Allegro.

X

X

'»•

Introd II Phrase

rVffff

Iff

^

•X— ^

i

44

THE EXTENSIONS AT THE BEGINNING OF A PHRASE. Par. JM,

6. AlUgro,

Introd.

B^

«=*=

A-X

II Pbrase.

1^

^Ih-^

trtj

7

I uJ I r"!"

r

Beethoven.

i

I

y^^

J ! ! ! J ! ^
^■M 1 J J

^ — 4> — 4» •^ -^i ^

a — # — ^ — a — ^ -• ' • — « — ^ 1 — ^" < ' 1 1 1 — ' — >-

7. Allegro,

BEETHOVEN.

i

i?

gSj: I y~x~l-t— *-H^

5^

i

■«5!-T-

-<5*-

Introd.

LET^"

i

Phrase.

I

^^-^

^^-*

:j:

E

^)y^jb— [-

t

l-l-

*i) In the first and second examples, the accompaniment precedes the
Phrase-melody by a few beats, without harmonic change.

*2) Here there is an actual chord-succession in the introduction ; but it is
derived from the accompaniment, and the entrance of the melody is conspicuous
enough to mark the beginning of the Phrase proper.

See also Mendelssohn, No. 35, first measure; No. i, first two measures;
No. 24, first two measures; No. 26, first three measures; No. 34, first two
measures ; No. 36, first three measures. The passage at the beginning of Nos.
19, 29, and 46 is too long to be called an introduction to the first Phrase. (See
par. 44.)

*3) The upper rests in this measure and the next one are only imaginary ;
while the introductory passage surely does not belong to the ** Melody", as
such, it is, nevertheless, the same tone-line, and leads uninterruptedly into the
latter.

*4) A tentative anticipation of the first figure of the Phrase-melody.

See also Beethoven, Symphony No. i. Finale, first ten measures ;
Brahms, Serenade in A, op. 16, Finale, first six measures ; Schubert, Pfte.
Sonata No. 6 (op. 147), Finale, first six measures.

Par. 29.

THE EXTENSIONS IN THE COURSE OF A PHRASE.

45

4. The Extensions in the Course of a Phrase.

29 > These represent probably the most important class of
Phrase-extensions, as they conduce more directly than those at the
•end or at the beginning, to the coherent development and growth
of the thematic substance of the Phrase as a whole. They consist :

(a) In one or more repetitions (exact or variated) of any welU
defined member of the Phrase, anterior to the Cadence-member.

It is manifest that all extensions of a Phrase will simply create confusion
and lead to a misapprehension of the writer's intention, unless they are recog-
nizable as '•''extensions " which merely add to the sum of heats and measures
without entirely destroying the outlines of the original {regular) Phrase^ or
cancelling the impression of a '* regular design" as fixed basis of the irregu-
larly inflated result. And, for this reason, this more misleading class of
extensions, in the course of a Phrase, can be safely undertaken only in those
phrases which contain strongly marked members or figures, with well-defined
extremities.

At any such well-defined " joint ", the Phrase may be — ^figuratively speak-
ing — pried apart to admit of the insertion of a repetition, sequence, or exten-
sion of any kind (and of almost any extent); and, if sufficiently marked in
character, the foregoing member will be readily recognized as the origin of
the interpolated reproduction.

For illustration :

X. Allegro.

-ExteDsion.

Haydn.

o9«

2. Andante.

-Extension.

46

THE EXTENSIONS IN THE COURSE OF A PHRASE. Par. 89b.

rExteusion.'-j

Haydn.

Haydn.

Allegro

*i) Repetition of first member (first half) of Phrase, in slightly modified
form. This class of repeated Phrase-members must not be confounded with
the illustrations of symmetrical phrase-formation given in Ex. 13, No. 2, and
Ex. 17. There they are essential^ and contribute to the 4-measure design;
while here they are unessentialy inasmuch as they constitute extra measures^
beyond the 4-measure design.

*2) This extension consists in a three-fold repetition of the first figure, in
alternating upper and lower registers.

*3) A sequence of the second figure, as in Ex. 40. If these two extensions
be cut out, a perfectly coherent 2-measure Phrase (the original design) will
remain. This is sometimes the case, but not necessarily so.

*4) A curiously modified version of the first Phrase-member. This exten-
sion, also, might be eliminated, leaving a regular 4-measure design.

(b) In one or more sequences (exact or modified) of any well-
defined member of the Phrase. For example :

X. Presto.

40. <

I

'h&

Exteusiou.

>*

-4 ^

-^-e-

rn^i

Haydn.

S^

^S

. 4^iJi^ i : i^ \ r '-r' I ff]

^m

/

Par. a9b. THK EXTENSIONS IN THE COURSE OF A PHRASE.

47

Mrndblssohn.

jJlSJIxnTU

%i vr^

i

3tz:

sion.

i

k

Haydn.

:T} fr

JXLfr> i

m

Ul^i -1-i.M^ ^^

^r=T^

m

$

4. A Ueg^o.
(Violin.)

I — Bxtension.

3

^^

f^

^E^^

^

* "y

*r#-

(i

£S

m=f

«6) I

(Orch.)

I

f

*-^

r I

^^^ ^ d y ^

rx

t — Extension

H 1 r

— 1 1 I I j^ I —

^zn — ft — I^

« J d ^ ^v^

8

t

8 8

T

Brahms.

48 THK EXTENSIONS IN THE COURSE OF A PHRASE. Par. 2Jte.

*i) A sequence of the second measure (figure) of the Phrase.

*2) A sequence of the second Phrase-member (or figure) in slightly modi-
fied form.

*3) A sequence of the first member.

*4) Sequence of the second member. As already intimated (Ex. 39, notes
*3) and *4)), these extensions are often so adjusted to the line of development,
that the members of the original 4-measure Phrase may be discovered in
coherent form without them. This, however, is by no means necessary ; it is
somewhat more likely that the interlined extensions will involve alterations of
the original melodic design ; or, better still, they will be so naturally inter-
woven with the texture of the Phrase as to affect, and even direct (appar-
ently) its current. But see the remark following 29a.

*5) Two sequences of the first member (five beats in length, and, therefore,
recurring in shifted measure).

*6) Two sequences pf the second member(three beats) in modified rhythmic
form.

The number of sequences or repetitions thus introduced in the
course of a Phrase depends upon circumstances, and can scarcely be
determined by rule. Smaller members or figures are usually repro-
duced a greater number of times than larger ones ; and sequences
may, without risk of monotony, be multiplied oftener than repeti-
tions.

More than one single exact repetition is hardly permissible ; in
the event of two repetitions, the last one (the third version of the
member) must be modified.

As a general rule, more than three successive sequences are
considered weak ; still, series of four (and even many more) se-
quences may be found in works of eminent rank ; especially when
consisting of smaller figures or members.

The student must appeal to his instinct and sense of proportion
and balance, and carefully avoid Monotony on the one hand and
Irregularity or Unevenness on the other.

(c) In the expansion of any prominent chord ; or of any im-
portant (prominent) melody-tone ; or melodic figure (of two or
three tones).

This is probably most frequently and most extensively applied
to the Tonic J chord (I J which almost always appears directly
before the two Perfect-cadence chords, usually on a strong accent
(see par. 6) . This special example of chord-expansion as extension
in the course of a simple Phrase, becomes the prototype of the more
or less elaborate *' Cadenza " in Concerto-movements. The conduct

Par. 89c.

THE EXTENSIONS IN THE COURSE OF A PHRASE.

49

of the upper parts (especially that of the melody proper) during
the prolongation of the chord, is largely optional ; but it is obvious
that it should te in strict thematic (or, at least, organic) agreement
with the rest of the Phrase, and not create the impression of a
foreign link. The rhythm may be treated with great freedom.

Frequently the bus stone alone of the I^ is prolonged as organ-
point, while the upper tone-lines sway about in optional har-
monic succession, either returning to the I^ or passing into the V
of the Cadence. In this case the I^ should appear on a strong accent.

But the expansion may also be applied to the Dominant chord
which precedes the final Tonic as first of the two Perfect-cadence
chords.

And it is furthermore applicable to any other chord, or chords,
in the Phrase, if they be of sufficient value and prominence to
warrant the expansion. For illustration :

I. Moderate,

iiV. ..J J I J ■ / .' ^F^^

J J

ereae.

Mendelssohn. No. 4.

•it -iii^-j ' j.

i

ff

i

I2 expanded.-

r

14 i t \ rrr\ i i j ji

*2)

^■■^

1

i-i-^i

I3 (expanded baBstone.).

50

THE EXTENSIONS IN THE COURSE OF A PHRASE. Par. 8»c

l^^f^M^T^^

J—;

Mbndblssohn. No. 4a.

-#— ^

^

'^='=f=i

«3)

^^

^

^

<^

t

-y — N

i^

y expanded.

T-

3. A ndanie.

}'Mi !

32:

^rfr

^^ ^

^=MdL

— m ^ j^

r-

r

Mendelssohn. No. 19.

ii^^^^^

i

#

4i

ri

^^

I

I

•6)

# ft-

g

^* ^« f^ ^

^

# ^ ft ^

S

I2 expanded.

Par. 89c. THE EXTENSIONS IN THE COURSE OF A PHRASE.

51

mjim

5- Larghet.'o,

^^^^ m$

Chopim. Op. 27 — T.

amr

6. Lento.

^^

Introd. —

r

■^-n^

T* Seguo.

^

^

r

\^

^jJ' f

Chopin. Op. 37—2.

7. ^ llcgro.

'y^^\^\> J ^;i J #^

(Original form.)
N.B.

E^^

N. B. Brahms. Sym. No. 3.

^m

-1^-

rfr l OTir l l

(Expanded form.)

*i) The expansion of the Tonic J chord enlarges the Phrase from its
original four measures to five.

*2) It is left to the student to discover how ingeniously this long expansion
of the Dominant-note in bass (as la) is interwoven, in the upper parts, with the
thematic structure of the 4-measure Phrase.

52 THE EXTENSIONS IN THE COURSE OF A PHRASE. Par. 29c.

*3) An expansion of the Cadence Dominant chord, which serves to coun-
terbalance the foregoing expansion.

*4) A long expansion of the Cadence Dominant chord. See also Men-
delssohn, "Songs without Words'*, No. 2, last ten measures; No. 33, last
five-and-one-half measures; No. 31, last eleven measures (particularly meas. 9-6
from the end); and also Ex. 59, meas. 11-12.

♦5) This exquisite example of Brahms (op. 79, No.i) contains several of
the above-mentioned extensions: at first, a sequence of the first Phrase-mem-
ber. At *6) an expansion of the Tonic J, with richly modified melodic
repetitions (three measures). At ♦y) the Cadence is due ; it is concealed by
continuing the rhythmic pulse in the inner part (imitating the Cadence-
member in soprano); and is followed by two repetitions of the Cadence-
member. At *8) the final Cadence-chord is reiterated.

*9) The expansion of this melody-tone extends the 4-measure Phrase to
five measures.

*io) The melody-tone a is expanded from a half-measure to a whole one.
See Ex. 62, note*3). A remarkable illustration of chord-expansion will be
found in Beethoven, Pfte. Sonata, op. 57, meas. 17-24 (beginning, as usual, to
count at the first complete measure).

See also: Chopin, Pfte. Sonata B|7-minor, Trio of Scherzo, first, second,
and third Phrases (expansion of final melody- tone) ; Chopin, Prel., op. 28, No.
2, first and second Phrases (ditto). Chopin, Prel., op. 28, No. 12, meas. 49-65
(expanded II', meas. 57-63). Also Ex. 42, meas. 3 and 4.

N. B. It is scarcely possible to overestimate the signiRcance
of ** Expansion" as a factor of composition. The principle of
expansion lies, in a certain broad sense, at the very root of the
entire art-growth, which carries along the germinal thought and
enlarges the raw harmonic material from one stage on into another
and higher stage, until the broadest development of the original
idea has been consummated. The larger (i.e., longer) forms which
are to follow the Phrase, will be found to be not altogether the
results of addition ^ but also of deduction and evolution ; not a
multiplication of the chords contained in the smaller design, but
these same chord-progressions expanded or magnified into broader
dimensions. Thus, the primary succession I, IV, V, I, which might
be the simple beats in a 4/4 measure, could become the four succes-
sive measures of a Phrase ; or they might be expanded into four
successive sections (at least in a general sense) of a still larger
design ; and so on. An illustration of this is found in Example 5,
in which the harmonic basis of the four measures is, approximately,
the very chords I, IV, V, !• suggested above. See also Example 10,
where the harmonic basis of eight measures is limited to no more than
three chord-elements; and Example ^a^ embracing only I, V, I ;

Par. 29d.

THE EXTENSIONS IN THE COURSE OF A PHRASE.

53

and Example 74, the eight measures of which contain no other
chord-progressions than I | V | I | V, I || V | V | V, I | IV, V, I||
(the last four approximated). This emphasizes the importance of
the advice given in par. 5, and in Exercise 2, '* to use as few chords
as possible".

(d) In the substitution of a new Cadence-member for the
original one, or the addition of a new Cadence-member, when
involved by foregoing extensions.

After such extensions, in the course of the Phrase, as may-
diverge widely from the original thematic line, or lead quite away
from the original (or expected) approach to the Cadence, it is often
necessary to invent an entirely new Cadence-member, Its extent
will depend upon where the ultimate cadential impression is required,
irrespective of the aggregate number of measures.

It is evident that, while new, it must, nevertheless, be strictly
coherent, and thematically homogeneous. For illustration :

X. Andante, i ^ i..^^^

u^^.-^^-::==^ — .-I i^i — 1

Frt r ||j J r

Mbnublssohn. No. x.

54

THE EXTENSIONS IN THE COURSE OF A PHRASE.

Par. 30.

3. Aiuiant$,

w=f^

^g^

Original form !| Repeated

SI

fc4:

*i) These two measures are an expansion of the chord c - e - g - b ; the
Soprano is thematically related to the c at the beginning of the 5th measure,
of which it is an anticipatory expansion. *2) The two half-notes are expanded
quarter-notes (half-measure expanded to whole measure). *3) The perfect
Cadence is due on this beat (see two measures later) ; it is evaded by a
Sequence of the foregoing measure. *4) The ** new Cadence-member". It
cannot be called a new ** Phrase *', not only because of its brevity, but because
no actual Cadence has preceded it, to terminate the current Phrase. It is "new",
however, inasmuch as it differs almost entirely from the original Cadence-
member (in meas. 5-6). *5) The digression begins with this cj{, which enhar-
monically represents the original d\}. The three following measures are
a repetition and expansion of the first melodic figure in measure 3.

*6) The " new Cadence-member " ; see note *4).

*7) A curious suppression of the Cadence-tonic in the highest part. Its
place is filled by the keynote in Bass.

30- The extension of a Phrase during its course, as above
shown, is rarely, if ever, undertaken at the very beginning of a
composition (i. e., with the Jirst Phrase or upon the Jirsi appear-

Par. 31. THE CHAIN-PHRASE. 55

ance of a Phrase). It is often very appropriately applied to the
repetition of a Phrase, as a means of modifying the second version
(see 19) ; and it may be introduced in any of the later phrases of a
section of the form. It is of peculiar value in the construction of
transitional passages and Phrases, as will be seen in due time.

BXBRCISB 6.

Former Phrases, or new ones, with extensions at the beginning (par. 28);

And in the course, according to 29 a and b. In the latter case the Phrase
is first to be written out (at least the Melody) in its original regular form, and
then extended (about as in Ex. 42, No. 2).

EXERCISE 7.

Extensions in the course, according to par. 29c and d. The Phrases are
first to be written out in regular form, and then extended.

CHAPTER IV.

THE CHAIN-PHRASE. MELODY-EXPANSION.
IRREGULAR PHRASE-FORMATION.

31 . The extensions in the course of the Phrase, particularly
the repetitions and sequences (par. 29 a and b) of small motives
and figures, are sometimes multiplied in such a manner and to such
a degree as to give rise to an irregular design, for which the term
*' Chain-phrase" may most appropriately be adopted, inasmuch as
the successive members assume the character of continuous small
links, in a chain of arbitrary form and extent. It can scarcely be
regarded as a legitimate design, for it is the type of formlessness,
and is, therefore, beset with dangers for the beginner. But, if
coherently developed, it is justifiable, and is often extremely effec-
tive in the service of necessary climaxes ; and also in transitional
sections, which tend persistently during many measures toward a
desired point, in uninterrupted career.

The impression of a Cadence must he scrupulously avoided^ as
it would separate the ** chain " of thematic links into a *' group"
of distinct Phrases.

56

THE CHAIN-PHRASE.

Par. 31.

Aivd a certain reasonable limit must, of course, be observed.
For illustration :

I. Alitgro vivace.

■Am

m

m

J 1 tf--/i-i — P f~

PP

♦1)

(^^

:f^^^

Vint

^^

^te

4< J I

i

^

eresc.

^»-tf#-y- i V f " ^ I iTT f I i f -^

rtt*

gii" rm I r

^1

fe

Beethoven. Op. 78.

«

i

.i?^ dim.

-^Ji

2. Agitato,

p pp etc.

g

^^

Phrase, repeated, and extended .

m

Par. as.

MKLODY-KXPANSION.

57

m

i^S

1 I

r

=n i

I I

J I

I I I

*2)

^^^

^

^*

:^

ft^

n

S -H

1^ : .^ ij ' * I

ore

seen

H fitt^F?^-M?^&4^

segae

r^-T

Mrndblssohn. No. xo.

*i) The original 4-measure Phrase extends to this beat. The thirteen
or fourteen measures which follow are all forged, like so many ** links ", out
of the second half of the Phrase, chiefly out of the figure contained in its third
measure.

*2) The " links** of theChain-phrase are here derived from the second and
third notes of the second measure of the original Phrase, and are reproduced
sequentially, at first in groups of three, and finally in progressive ascending
succession.

*3) This measure corresponds exactly to the original second measure.

*4) This example embraces no less than twenty measures (including the
original Phrase and its repetition), all developed out of a brief 4-measure
Phrase, and so interlocked as to avoid a cadential impression at any point in the
course of the Chain-phrase.

See also Ex. 56, first thirteen measures. Ex. 85, meas. 9-19.

See also: Mendelssohn, ** Songs without Words**, No. 2, meas. 35-20
from the end; No. 14, last ten measures; No. 17, end of meas. 28 to meas. 35;
JVo, ^Sf meas, 21-10 from the end; No. 45, meas. 24-12 from the end.
Beethoven, Pfte. Sonata, op. 31, No. 3, first movement, meas. 34-19 from
the end. Beethoven, Pfte. Sonata, op. 28, first movement, meas. 57-94
from the double-bar (thirty-eight measures long, and preceded by a similar
Chain-phrase of sixteen measures). Schubert, Pfte. Sonata No. 7, (op. 164)
first movement, meas. 16-27; Finale, meas. 71-94. Brahms, op. 119, No. 3,
last 22 measures.

Mklody-kxpansion.

32. A simple brief Phrase, or Phrase-member, may be utilized
as a thematic germ, and be developed, by means of repetition and
tone-expansion^ into a melodic thread of considerable length. This

58

MELODY-EXPANSION.

Par.ftl.

process differs from that employed for the Chain-phrase, in that,
as a rule, it does not involve the " sequence ", nor the necessity of
avoiding slight cadence-impressions in the course of the entire
expansion. It is most commonly, though by no means exclusively,
encountered in the closing sections (Codettas, see par. 51) of larger
forms, where the gradual enlargement of a foregoing thematic
member accords well with the gradual relaxation of the rhythmic
and harmonic momentum. It is cited here more by way of illus-
trating a resource of great future value, than for special exercise by
the student, at this point. For example :

X. Andante,

inn

t

Original form.

^

etc.

t

X

t

Exjumded form ( at end of

I

b.

MSNDBLSSOHN. No. 37.
♦1)

%

t

m

i

eompodtioii).

til Ik *• Allegro.

R R R fq IT"

Original form.

Repetition and Expansion

y

Sv^

-^ fT I nn fr~i w\ irn

fe^^^

Mbndblssohn. No. 32.

X

m

■jSL

■^-

^

^=¥^

X

3. A ndante.

LJ(SZ2.

i

a.

5

t

Original form.

etc.

Expanded form.

Par. 83.

IRREGULAR PHRASE-FORMATION.

59

OrigJDAl form.

i

y

I

t

E5

tiM:

Expanded form.

i

^

# 4 #

Mbndelssohn. No, 24.

3tZ±

liizf:

^

a.

m

^

i

etc.

5. A ttdaMte.

i

r

a.

i

E3

1=^

I

b.

i3=t

etc.

t

Expanded form.

Original form.

i

I I Beqaence of a. | | sequence of b.

^FS

P

^

t

^

J*

'^'- -^

ra.-:;

^t

ji

£

I

^

Brahms. *3}

^

c.

etc.

*i) The pupil is expected to refer to each ** Song without Words" indicated
by number, and to examine, carefully, the entire harmonic equipment (which
is of very great importance), and the general surroundings. The extracts
given will all be easily found, near the end of the number.

*2) These three tones are apparently a gratuitous insertion, but they
relate, as a kind of sequential anticipation, to the figure marked ** a ".

*3) Third Symphony, andante movement, near the end. The relation of
figure ** c " to figure " a ** must not be overlooked.

Irregular Phrase-formation.

1. It is sufficiently obvious that the modes of extension and
development explained in par. 24-32 are liable to impart an irregular
metric design to the original regular 4-measure Phrase (as regards
the aggregate number of measures).

6o

IRREGULAR PHRASE-FORMATION,

Par. 83.

But it is also possible that the original conception of a Phrase
may assume an irregular dimension, namely : a length of 3, 5, 6, 7,
9 measures ; and examples of this kind are by no means rare.

The irregularity results either from reaching the Cadence a
measure too soon, and then simply omitting the following (legiti-
mate Cadence-) measure, in order to relieve the Cadence of unneces-
sary weight; this accounts for the majority of 3-mea8ure and 7-
measure Phrases ;

Or the irregularity results from an unequal association of
melodic members. For illustration :

X. A ndante.

4S. -

S-measnre Phrase. *1)

'wfr^^r^'^^su^nPrir'^ ^

Mendelssohn. No. 40.

P=^

cvad. Kepetitioo.

2. A llegro»

t

1^

\

nr^^^

Mozart.

3-measure Phrase. *2}
3. Adagio,

Sequeuce.

Schubert.

w

fe

-4- — #

-^-

5-measare Phrase. *3)
4. Largo,

S

^

■^9-

§

Beethoven.

S-measare Phrase. *4)

5. Allegro,

i

t

^igrp

^ — — I 1 1 I r P #

Haydn.

^ I '^ri^uur

6-measnre Phrase. *5)
6. AlUgreito.

•^p—S-

bSEfczfc

t

^j-"-* — *— * — *-

e^measare Phtase. *6)

i^-ijlj j'j^Tj ^^

Par. 83.

IRREGULAR PHRASE-FORMATION.

6l

Schubert.

7, AlUgrttto,

%^:-^\i\i^* \ \ ^^

t

^

t=c

5

J.-r

-i.

^fl\lium^

T-measnie Phrase. *7)

Mbndblssohn. No. xa

I*

^i

*i) This is a somewhat unusual specimen of a Phrase of three large
measures in sloiv tempo. The 3-measure Phrase is quite frequently merely a
magnified example of the triple- ;«^aj« re, each beat corresponding to an entire
measure ; and when this is the case it is most common and most intelligible in
rapid tempOy and generally appears in groups of at least two or more Phrases.
See Beethoven, Symphony No. 9, second movement, twenty-seven meas-
ures from the first repeat {ritmo di ire battute). Mendelssohn, "Songs without
Words ", No. 9, first three measures ; No. 16, first three measures. Schubert,
** Winterreise '% No. 18, first three measures.

*2) 3-measure Phrase in more animated rhythm, and followed by a se-
quence. Compare Note *i).

*3) This 5-measure Phrase results from the union of one 2-measure with
one 3-measure member.

*4) The 5-measure dimension might be accounted for, in this case, as the
result of a modified repetition of the first measure, — i. e., regarding the second
measure as superfluous; compare with Ex. 39, note *i), in which there is no
doubt about the unessential nature of the extra measures. It appears probable,
however, that this second measure is essential. See also Beethoven, Pfte.
Sonata, op. 27, No. 2, first five measures. Schubert, "Winterreise '% first
five measures of Nos. 15, 19, 20.

*5) A6-measure Phrase, consisting of three members of two measures each.
See " God Save the King ", first six measures.

*6) Two members of three measures each. See also Beethoven, Pfte.
Sonata, op. 26, last movement, first six measures.

*7) This Phrase aggregates only seven measures, simply because the
Cadence-tonic falls prematurely (in the seventh measure) and is not held
throughout the eighth measure. Compare with meas. 15-22 of the same
" Song without Words " ; and see ** Song without Words *', No. 6, first seven
measures.

MISCELLANEOUS EXAMPLES OF PHRASE-EXTENSION.

Beethoven, Pfte. Sonata, op. 31, No. i. Adagio^ last twelve measures
(2-measure Phrase).

Beethoven, Pfte. Sonata, op. 28, Finale, last eighteen measures (4-meas-
ure Phrase).

Beethoven, Pfte. Sonata, op. 28, Finale^ meas. 29-43 /Phrase repeated
and extended).

Beethoven, Pfte. Sonata, op. 2, No. 2, Largo^ meas. 13-19.

62 THE PERIOD-FORM. Par. 80.

Haydn, Pfte. Sonata, No. 14 (Cotta ed.), Adagio^ last fifteen measures.
** " ** No. 17 ( ** " ), first eight-and-one-quarter meas-

ures.

Beethoven, Pfte. Sonata, op. 10, No. 3, Largo^ meas. 17-26 (Phrase re-
peated and extended).

Beethoven, Pfte. Sonata, op. 14, No. 2, Finale^ last eighteen measures.
** " " op. 26, /7«fl/«, meas. 321^-48 (4-measure Phrase,

each ^a//" several times repeated).

Beethoven, Pfte. Sonata, op. loi, Finale^ last fifteen measures.

BXBRCISB 8.

Former Phrases of 4-measures, with a rej>etition expanded into Chain-
phrase form. A few experiments in the process of Melodj-expansion and in
Irregular Phrase-formation may also be made.

CHAPTER V.

THE PERIOD-FORM.

34. The Period consists in the union of two phrases,
extending consequently, when regular, through eight ordinary
measures in ordinary moderate tempo. But see par. i.

35. The first of the two Phrases, called the Antecedent
Phrase, begins with the Tonic chord, as a rule ;

But it does not end with the perfect Cadence like the simple
Phrase, because such a Cadence would completely finish the Phrase,
and render the addition of a companion-phrase, in coherent (un-
broken) succession, not only impracticable but unnecessary. (See
par. 23.)

The Cadence of the Antecedent Phrase must, therefore, be
made in such a manner as only partially to check the harmonic and
melodic current. Such comparatively lighter interruptions are
called Sbmicadbncbs.

36. The harmonic form of the Semicadence may be best
defined negatively, as *'any chord-association (with cadential
effect) which is not the perfect Cadence " (see par. 3).

Par. 87. THE PERIOD-FORM. 63

The most common and natural Semicadence for an Antecedent
Phrase is obtained by resting, at the prescribed accent, upon some
Dominant chord, usually as concord, but occasionally in one of its
discord-forms ; this Cadence-dominant may be approached through
tf«y convenient chord (see Ex. 46, Nos. i and 3; Ex. 47, Nos. 2
and 3).

But the Semicadence may also fall upon a Tonic chord, on
condition that its root does not appear in either outer part in such
prominence as to suggest the perfect Cadence (especially when the
Tonic is preceded by a Dominant chord) ;

Or, more rarely, upon the triad VI, the Relative of the I (see
Ex. 48, Nos. 2 and 3; Ex. 60).

Or it may fall upon a Subdominant chord (IV or II), though
this is rarely the case (Ex. 46, No. 2).

Or the Antecedent Phrase may modulate^ so as to end upon
some chord (usually the I) of a next-related key, in the following
order of preference :

Upon the I of the Dominant key ;

Upon the I or V of the Relative key ;

Upon the I of the Dominant-relative key ;

Upon the I (rarely the V) of the Subdominant-relative key ;

More rarely upon the I of the Subdominant key. (Ex. 47,
No. I ; Ex. 48, No. i ; Ex. 59; Ex. 68, — ^first Cadence of each.)

As regards the location and rhythmic treatment of the Semi-
cadence, see par. 3 and 4; and Ex. 46, note *i).

3T. The harmonic and melodic character of an Antecedent
Phrase, while necessarily conforming to the directions given in
paragraphs 5, 8-12, will naturally be influenced to a certain extent
by the altered condition of the Cadence ; and, as a rule, the choice
of Semicadence-chord should be more or less definitely determined
beforehand, as the aim toward which the current of the Phrase
will tend.

It is at least certain that the less decisive nature of a semi-
cadence imparts a corresponding unfinished (conditional or inter-
rogative) character to the Antecedent Phrase, and renders it
dependent upon its following companion for completion and
counterbalance. It is a thesis^ dependent upon its antithesis ; a
Question ^ awaiting its Answer,

64 THK PARALLEL CONSTRUCTION. Par. 39a.

38 • The second of the two Phrases in the Period-form is
called the Consbqubnt Phrase. It begins with any chord, or
key, which is reasonably adjustable to the preceding semicadence-
chord ; and it ends with the perfect Cadence, in the manner de-
scribed in paragraphs 3, 4, 6.

Viewed in its metric aspect, the Consequent Phrase is, w^hen
regular, exactly similar in length to its Antecedent ; and both its
beginning and its cadence usually correspond (as to their location
in the measure) to beginning and cadence of the Antecedent Phrase.

39. Viewed in its Mklodic aspect, the Consequent Phrase
must, obviously, preserve close relationship w^ith its Antecedent,
though absolute corroboration is not obligatory.

The Period-form maj (and doubtless should) be regarded as a magnified
Phrase ^ whereby the "figures" have grown into "members", and the
"members" into complete "Phrases"; and the "quarter-cadence" (par. 8)
between the phrase-members has developed into a " semicadence". From
this point of view, the metric conditions illustrated in Examples 12 and 13,
and emphasized in paragraph 12, also apply, though in a broader sense, to the
simple Period-form.

The following distinction^ influence, and possibly govern, the
melodic structure of the Consequent Phrase :

(a) The Parallel Construction.

In this form, the melody of the Consequent-phrase more or less
exactly corresponds to that of the Antecedent ; sometimes so closely
that it is prevented from being actual Phrase-repetition (the form
out of which the ** parallel Period " is evolved) only by the neces-
sary distinction in the formation of the two Cadences, namely, the
Semicadence of the Antecedent, and the perfect Cadence of the
Consequent, which, of course, must essentially differ from each
other, in order to constitute the Period-relation. Compare para-
graph 21. This is the primary and most natural variety of the
Period-designs.

Usually, however, the similarity between Antecedent and Con-
sequent is less pronounced, and often only a general resemblance
exists.

In some cases the parallelism is established by constructing the
Consequent Phrase as sequence (partial or entire) of the Ante-
cedent. For illustration :

Par. 39a.

THE PARALLEL CONSTRUCTION.

65

z. Alligretto^

^4n '■>' f

i

tt

t r T •

t ^- 1 ^

rj \ r

46.

Antecedent Pbnwe.

^^

^3

Brbthovsn.

'■^ ^'i ,K ^ Li ^

^

Perfect Cad.

*2)

^

^

5 5

1 I I I i-T-r

^ !-

i

a. Allegretto.

Antecedent Phntae.

*8)

AblV

66

THE PARALLEL CONSTRUCTION.

Par. 30a.

8va

h

^A^j^ i::fj^ f7^

p

m

segae

Chopin. Op. 47.

Perf. Cad.

I^>.

H*

^

^^

^

Semi-

Antecedent Phrase.

Folk-song.

P i J J J 1 1 1. /J H i ^^Jlri i -J-

I

Conseqnent Phrase.

^

^

r

^^pi

|V 1

*i) The Semicadence rests upon the Dominant concord, which falls upon
the first beat of the Cadence-measure, and is projected into the second beat.
The rhythmic pulse is continued, exactly in conformity with the principles
which govern the ** concealed cadence" (par. 18, which see). The necessity
of thus ** bridging over" the cadence-space between the two Phrases, is owing
chiefly to the momentum of th<* accompanying figuration — a factor which
never ceases its motion during any transient cadence, and usually impels itself
even beyond the final perfect Cadence. But it is generally desirable, even in
the absence of a figural part, to connect the Phrases as closely as is compatible
with a perfectly unmistakable cadence-impression. After this latter condition
has been fulfilled, it is not even necessary to remain any longer within the
semicadence-chord. This is illustrated by the notes in small type which
occupy the last fractions of the 2nd beat.

Par. 89b. THE OPPOSITE CONSTRUCTION. 67

*2) The Consequent Phrase is a literal reproduction of the Antecedent,
only excepting the very last chord, which, being the Tonic, constitutes a per-
fect cadence, and establishes the indispensable condition of opposition to the
Antecedent Phrase. See again par. 21.

*3) This Period begins upon the Subdominant harmony, in consequence
of occurring in the course of a larger composition.

*4) The Semicadence also rests upon the Subdominant, and, therefore,
l«ads over very smoothly into the Consequent Phrase, which is identical with
the Antecedent (excepting changes in register and rhythm) up to the penulti-
mate chord.

*5) The Semicadence rests upon the Dominant harmony, and the rhythm
is completely checked, as is most customary and appropriate in music of such
a simple character. For further exaniples of the Period of preponderantly
parallel construction, see Mendelssohn, " Songs Without Words,'* No. 28,
meas. 5-12, and meas. 13-20 (imperfect Cadence) ; No. 29, meas. 5-12; No. 35,
meas. 6-13; No. 22, meas. 10-17 (imperfect Cadence). Ex. 49 (of this book),
first four measures; Ex. 55; Ex. 72, first eight measures; Ex. 73, meas. 1-8;
Ex. 75, meas. 1-8; Ex. 84, meas. 1-8; Ex. 85, meas. 1-8.

Furthermore, Beethoven, Rondo, op. 51, No. i, meas. 1-8;

Beethoven, Pfte. Sonata, op. 27, No. i. Finale, meas. 1-8;
Beethoven, Pfte. Sonata, op. 10, No. 3, Menuetto, meas. 1-16 ;
Brahms, Symphony I, last movement {Allegro), meas. 1-8.
Jchumann, Jugend-Album, op. 68, No. 23 ("ReiterstUck*'),
first 8 measures; prevented from being a "repeated
Phrase '* by very last tone only.

(b) The Opposite Construction.

In this variety of the Period- form, the melody of the Conse-
quent Phrase pursues largely or entirely the direction opposite to
the melodic progressions of the Antecedent.

It can scarcely be claimed that this is a natural or spontaneous mode of
construction, for it is based upon a process which is largely mechanical, and
may result, when applied too strictly, in constrained and unmelodious tone-
lines. But it represents the same resource of thematic development as the
*^^ Imitation in Contrary motion " in the Polyphonic forms, and is of value to
the student because of its suggestiveness. In Homophonic composition its
occurrence is rare.

Occasionally the delineation of the Antecedent Phrase is almost,
or quite, literally thus inverted in the Consequent ; but usually only
the general melodic drift, or a few of the more striking members,
reappear in the opposite direction. For example :

68

THE OPPOSITE CONSTRUCTION.

Par. 80b.

I. AlUgrot

47.

GOBTSCHIUS.

a« Moderato.

Semi-cad. W *3)

Anteoedent

^^M^

Conseqaent.

Haydn.

3. A ndanU.

. 5

Antecedent.

Semi-cad.

:r— llj — V Con-

Mbndblssohn, No. 37.

sequent.

*i) The Semicadence rests upon the I of the Dominant key (with major
3rd, 1. e., the " modulatory Stride "). The Consequent Phrase-melody is (in-
tentionally) the literal contrary direction of the Antecedent melody, from
beginning to end.

Par. 89c.

THE CONTRASTING CONSTRUCTION.

69

*2) This is by far the best and most common of all the forms of the Semi-
cadence, viz. : the V, preceded by the accented I m J inversion. The latter
(the I9) precedes the actual Cadence-chord (V) as embellishment (** appog-
giatura"), and fulfils chiefly the important purpose of "bridging over** the
cadence-space (Ex. 46, note *i) ).

*3) The opposite direction of the Antecedent melody is maintained
through about one-half of the Consequent Phrase, only excepting the prelim-
inary tones.

*4) Here ihe general drift of the Antecedent (more especially of the ac-
companying harmonies) is reversed in the Consequent. This is also the case
in Mendelssohn, No. 45, first eight measures. See also Ex. 54.

See also : Hummel, Pfte. Sonata, op. 13, Adagio^ first eight measures ; and
Finale^ first eight measures.

(c) The Contrasting Construction.

This variety of the Period-form is characterized by the absence
of such parallel traits (in the same, or in the opposite, direction) as
those explained above. The principle of *' contrast" prevails,
though not necessarily to the utter exclusion of parallel figures.

As has already been intimated, these three distinctions are to be accepted
and applied more in a general sense, than in strict detail. The Consequent
Phrase of a Period may reveal traces of all three varieties of construction,
without belonging distinctly to either class; and probably the majority of
Periods will be defined as "contrasting** construction, merely because suffi-
cient evidence of parallel or opposite construction is wanting.

The *' contrast" is generally limited to the melody (i. e.,the
THEMATIC element), and is rarely extended to the style, rhythm
and general character (i. e,, the formative elements).

Hence the rule, for comparatively small forms, that while
thematic contrast is permissible and necessary to a certain degree,
formative relation must be preserved. For illustration :

I. MotUraio*

H->H

Antecedent.

£

m

rfc

w

t

-i^— r-

•1)

SCHUBBRT.

Conseqaent. '

70

VARIETY AND UNITY.

Par. 40.

3. Vivaee,

•2)

F4 .r ■^ ^•\r^ l ^^M

"J J-MV J JJtiJ

Antecedent

ConaeqnenL

Haydn.

Andantino.

± t: ± t:

r V T V

segae

1^

Conso-

Chopiw. Op. 38.

m

m

^

qnent.

*i The Semicadence rests upon the Dominant of the relative key, and
obtains its cadential effect through the purely artificial pause (/^), which
marks the otherwise vague point of rest. There is not a single point of
melodic resemblance between the two Phrases. And the Consequent begins
(unlike the Antecedent )upon a preliminary beat.

*2) Semicadence rests upon the Tonic triad, but with its Fifth in Soprano.
The contrast between the two Phrases is here unusually marked.

*3) Tonic Semicadence, with Third in Soprano. The contrast is distinct,
though traces of parallelism exist.

See also Ex. 44, No. 2 (probably a 4-measure Period) ; Ex. 52 ; Ex. 53,
meas. 1-8; Ex. 60, meas. 1-8; Ex. 61, meas. 1-8; Ex. 71, each Part; Ex. 89,
meas. 1-8. Msndelssohn, "Songs without Words", No. 9, meas. 3 (2d
half )-7 ; No. 48, meas. 1-4. Beethovbn, Symphony IV, Adagio^ meas. 2-9.

Variety and Unity.

40. Among the most vital and ever urgent conditions and
demands of perfect artistic creation are those oiF Variety and
Unity, — variety in unity, and unity in variety. Though the con-
dition of unity appeals first in order to our consciousness, that of
variety is of at least equal (ultimately of even greater) importance ;

Par. 40. VARIETY AND UNITY. 7I

and each asserts its claims with equal emphasis. These two
opposing forces are therefore incessantly at work in every creative
process, and the skill of the master is displayed in so directing
them that the most just balance is preserved between them.

The condition of Unity dictates all the methods by which
fundamental regularity, concentration, agreement and corrobora-
tion are obtained ; these methods are :

The metrical arrangement of equal beats in equal measures
and sums of measures ;

The adherence to a central tonality ;

The usual agreement of component members (£xs. 17 and 18;
par. 38, 39« and d) ;

All repetitions, sequences and imitations (par. 19).

The condition of Variety dictates all the methods by which
this fundamental regularity is divested of its monotony and
fatiguing effect, and by which a judicious degree of grace and
contrast may be gained ; these methods are :

The modification of the rhythmic measure by use of longer
and shorter tones ;

The modulatory movements around the tonal center;

The alterations of register (par. 19^)1 — of style (i9</), — of
dynamic effect, within the same regular harmonic limits ;

The sequence itself (as distinguished from exact repetition) ;

All unessential modifications (Ex. 19; Exs. 26-30), which
sustain interest without jeopardizing the condition of unity, by too
effectually disguising the necessary thematic and formative agree-
ment.

The impossibility of formulating accurate rules for obtaining this *' Variety
in Unity** is clearly manifest. It is simply to be dome in mind by the
student, and will thus exert a moulding influence upon his musical views and
habits.

EXERCISE 9.

Invent a number of Periods, of diversified character and design, chiefly
8 measures in length. Alternate regularly between the major and the minor
modes ; employ the different varieties of duple and triple time, and the differ-
ent grades of tempo (from Adagio to Presto), Exemplify the various forms
of construction (parallel, opposite, contrasting).

Work toward the Cadence during each Phrase, and from Cadence to
Cadence. Make both Semicadence and perfect Cadence suflficiently distinct •,
avoid, as beginner, vague forms of all cadences; all concealing or bridging-

72

THE REPETITION OF THE ENTIRE PERIOD-FORM.

Par. 4L

over of the Semicadence must be effected with strict consideration of a per-
fecily definite cadential impression.

Endeavor to preserve distinct melodic character; let there be a strongs
prominent and continuous melodic line in some part or other (chiefly in
Soprano) always.

In other words, the "tune'* ("air'*, "cantilena*') should always be
conspicuously the dominating purpose, no matter what the style or tempo is.

Next in importance to this is the leading of the Bass part <i which should
be as smooth and melodious as possible, and should be sketched against the
Melody alone before the inner parts are added. See also the n. b. on page 52.

CHAPTER VI.

THE DEVELOPMENT OR EXTENSION OF THE

PERIOD-FORM.

4 1 . The modes of extension explained in Chapter III are
applied to the Period in the same general manner as to the Phrase,
though on a correspondingly broader scale.

As usual, the principal factor is that of repetition^ which may
be applied to the entire Period bodily, or to either or both of its
Phrases separately.

I. The repetition of the entire Period-form.

<

The Period may be repeated literally^ without any other
changes than those involved by concealing or bridging-over the
perfect Cadence, before repetition. See pars. 18 and 21.

But here again it is far more usual and desirable to introduce
unessential modifications^ during the repetition, corresponding to
the modes of manipulation described in paragraph ipdf, b^ c, d^ e —
which review. For illustration :

Allegro.

V t:t^£

Par. 42.

THE REPETITION OF THE CONSEQIJENT PHRASE.

73

^ -^ • -^ II Repetition.

n. ^ '^ i J J ^ ',i ^ ^ i^j-H^ ^i

*i) Neither Semicadence nor perfect Cadence requires any bridging-
over, in this instance, because the " space '* is reduced by the beginning of
each Phrase three beats in advance of the primary accent.

*2) The modifications during the repetition consist in shifting the melody
down an octave, and altering the style of the accompaniment ; (also in the
dynamic change 'from // to/; and a change in orchestration — see later).

See also: Mendelssohn, "Songs without Words", No. 9, meas. 3 (2nd
half)-ii; No. 27, meas. 5-20; No. 29, meas. 4 (2nd half)-2i, — one measure of
interlude between Period and its repetition, as in Ex. 31.

Beethoven, Symphony IV, Adagio^ meas. 2-17 ; Pfte. Sonata, op. 13,
Adagio^ meas. 1-16. Pfte. Sonata, op. 28, Scherzo^ first thirty-two measures
(large i6-measure Period, repeated).

Chopin, Prel. op. 28, No. 19, meas. 23-7 from the end. Prel. op. 28, No.
21, meas. 17-32.

Schubert, Impromptu, op. 90, No. 2, first twenty-four measures (two

«

repetitions) .

2. The REPETITION OF THE CONSEQUENT PhRASE.

42. In case eitlier of the two Phrases of the Period-form is
to be repeated alone^ it is much more likely to be the Consequent
than the Antecedent. This corresponds to the repetition of the
second half of a Phrase, defined in paragraph 2^a,

The repetition may be, as usual, exact or modified, — generally
the latter (par. 19).

The treatment of the perfect Cadence, before the repetition
of the Consequent, must conform precisely to the conditions

74

THE REPETITION OF THE CONSEQJJENT PHRASE.

Par. 42.

explained in i8 and 21, which review; i. e., the Cadence may be
bridged over, but must not be completely evaded^ by aay essential
change of harmony. For illustration :

50.

Adagio molto.

Consequent

Bepetition.

«2)

Bbbthovbn.

^^;^C^^m

■i i 1 i

r?

E

I

■1^-*:

\

*i) The perfect Cadence, due at this accent, is concealed by passing on
chromatically in Soprano into the chord-Third ; and the space is bridged over
by continuing the chromatic succession directly into the first tone of the
(repeated) Consequent.

*2) Compare this measure with measure 7, and observe how the very same
melodic figure is here manipulated, in order to obtain an earlier and stronger
cadence-impression.

See also: Mendelssohn, ** Songs without Words ", No. 14, meas. 17-28
from first double-bar; No. 19, meas. 24 (2nd half)-30.

Beethoven, Pfte. Sonata, op. 2, No. 3, first twelve (13) measures (slightly
expanded cadence).

Chopin, Prel. op. 28, No. 20, entire.

Par. 43.

THE REPETITION OF THE ANTECEDENT PHRASE.

75

3. The REPETITION OF THE ANTECEDENT PhRASE, OR OF

BOTH Antecedent and Conseqjjent.

43. As already stated, the repeated Antecedent is of far more
rare occurrence than the repeated Consequent. It is most likely to
appear in connection with the latter, i. e., each Phrase separately
repeated.

The treatment of the Cadences must conform, here again, to
the rules given in paragraph 21. For example :

X. Moderato.

m

^

s

V

Repetition of

*^

^

81. <

J. :^ ±t 4 J i i

'^

i

m

J.

m

F V

m

Antecedent.

i

V

t

Oonseqneut.

i

^

f

r^f ' r ^

Sf-

t

<

2.

1SZ

J J Ji

*1)

J.

m

A.

i

f

ISL

f

t

I, V

^m

Repetition of Consequent.

f

r ' r

Mozart.

<

te

m

i J J

I
*2)

f=t

^

«3)

A

P

E

7

J . J J . ^i

=f=!l f f -^^=¥

^

§

76

THE REPETITION OF THE ANTECEDENT PHRASE.

Par. 48.

3. Presto.

^^^

X • Li ^ ^ X

y Bepetitioii of Antecedent

» * #

^^

Aocompaniment simile .... I

Coneeqnent.

V

•4) I

-1 1 iij. j. J. J.

►6)

Repetition of Consequent.

Mbndelssohn. No. ai.

*i) The Semicadence, here, corresponds exactly to the first one, four
measures back.

*2) The perfect Cadence is slightly modified (concealed) by placing the
chord-third in Soprano. Otherwise it agrees with its repetition, four measures
later.

*3) The Melody is considerably changed, during this repetition, though
not essentially so. It belongs partly, for that reason, to the ** Group "-forms,
explained in paragraphs 54-55.

*4) The Antecedent and its repetition close with precisely the same Semi-
cadence (on the Tonic chord). It will be observed that this example begins
with a Dominant chord ; that is because it occurs in the course of a larger
form. The same is true of the preceding illustration.

Par. 44. BXTENSI6NS AT THE BEGINNING OF EITHER PHRASE. 77

*5) The Consequent Phrase contains eight measures (twice the length of
its Antecedent), owing to chain-phrase extension (par. 31).

*6)6) The harmonic alterations during the repetition (a persistent inclina-
tion into the Subdominant keys) are noteworthy. See the original (** Songs
Tvithout Words'*, No. 21, thirty-eight measures from the end).

Mozart, Pfte. Sonata, No. 11 (Cotta ed.), first movement, meas. 29-8 from
the end. Bkkthovkn, String-quartet, op. i8, No. 3, " Minore " of third move-
ment, first twelve measures.

BXBRCISB 10.

Former Periods, or new ones, with complete repetition (according to par.
41). The repetitions must be variated by means of unessential modifications,
care being taken, however, to preserve the original harmonic form of the
Cadences.

BXBRCISB //.

The same, with repetition of the Consequent Phrase (par. 42).
The same, with repetition of the Antecedent Phrase, or both Antecedent
and Consequent Phrases (par. 43).

The repetitions must, here again, be unessentially modified.

4. The EXTENSIONS AT THE BEGINNING OF EITHER PhRASE,

OR OF BOTH Phrases.

4*4*. An extension at the beginning is, naturally, more likely
to be applied to the Antecedent Phrase than to the Consequent.

As usual, it will assume the character of an Introduction, as
explained in paragraph 28, which review. But in serving, as it
may, as an introduction to the entire Period, it can be somewhat
longer, and, at the same time, a trifle more individual in character
than the introduction to a single Phrase.

This is the case, to a moderate degree, with the passage at the beginning
of Mendelssohn's '* Songs without Words", No. 29, and No. 46; and, more
strikingly, in No. 12, No. 15, and No. 32. In each of the last three of these
examples, the preliminary passage might be called an " Introductory Phrase" ;
for, on account of its length and importance, it appears to fulfil a more inde-
pendent mission than that of an extension at the beginning of the first Period
only, and refers, properly, to the whole piece. But the style is, nevertheless, kept
so subordinate that no doubts can arise as to the actual beginning of the Period-
melody ; and, moreover, it is inseparably connected with its Period by ending
upon the Dominant harmony.

See also: Chopin, Mazurkas Nos. 9, 34, 13, 15, 17, 20, 21; more pro-
nounced, Nos. 7, 31, 2.

78

EXTENSIONS AT THE BEGINNING OF EITHER* PHRASE. Par. 46.

45. When such an introductory passage is conducted into a
complete Tonic Cadence, and thus separated from the following
Period, it ceases to be an ** Introduction ", in the strict sense of the
term, and becomes a

''Prelude''.

The moment this distinction arises, — in consequence of the separat-
ing effect of the perfect Cadence, — the necessity of preserving close
thematic and organic relation with the following Period is can-
celled, and complete independence of character (in everything
excepting tonality, meter and tempo) may be imparted to the
Prelude,

Comparatively close organic relation with the following Period exists in
the Preludes to ** Songs without Words " No. 19 (comp. Song No. i); No. 6
(first seven measures); No. 21 (first eight measures); and even in Nos. 3, 28
and 41, though these are severed from the first Period bj a double-bar. See
also: Chopin, Mazurkas Nos. 3, 42, 46.

Entire independence is illustrated in the Preludes to ** Songs without
Words" No. 4, No. 35, No. 16 and No. 9. The opening (4) measures of No.
27 d^fy exact classification ; in the independence of their character they sug-
gest the Prelude-class, but they lead, like an Introduction, directly and
smoothly into the first Period, from which only a Dominant Semicadence
separates them.

46- It is of course possible to interline a brief '* introduc-
tion ' ' to the Consequent Phrase also, but examples of this kind
are very rare^ because of the attendant danger of destroying the
necessary continuity between th6 two Phrases. Such an introduc-
tion to the Consequent will sound most plausible when correspond-
ing in thematic character and style to an introduction to its Ante-
cedent Phrase ; and it must be so skilfully handled as not to interfere
with the impression intended to be conveyed by the original (un-
extended) Period. For example :

52. <

Vivace^ non trop^o.

Voice. *1)

lutrod.

Antecedent Phrase.

Par. 47. THE EXTENSIONS AT THE END OF EITHER PHRASE* 79

t=^

W=t^

V — t

r r. ^ \ jTn^ ^

g:^^ f: #^ f

Iiitrod.<

i

5

iif

*

r ^ r/^-H'

X < l

<

(^^

9i

^

1^

^ J cre«c. J

tr

-y-^

d

*:

ItfL

t

Consequent Phrase.

Schubert.

*i) The vocal melody of this example (" Winterreise ", No. 2) fixes the
design of the ** Period ", beyond the possibility of misconception.

5. The EXTENSIONS AT THE EnD OF EITHER PhRASE, OR

OF BOTH Phrases.

4T. The extensions at the end are much more likely to be
applied to the Cadence of the Consequent Phrase than to that of
the Antecedent. They conform exactly to those explained in para-
graphs 25, 26, which review. One peculiarly instructive illustra-
tion will suffice :

Mtnueiio.

83. =fcf

i

iz

m

,us

^

±

^

atrjt

-•-arl-

-^-y-g

segne

8o

THE EXTENSIONS AT THE END OF EITHER PHRASE. Par. 47.

^1 ^^- %? ^

Y J 3

s

^ f:

Consequent.

^

»E

ES3

S

K

j- X X

*1) «2)

'• •

:^

P *3)

M^

r ^ hriri^

1

2 expanded (Ex. 41.)

Haydn.

^^

S?^ I ^

I

FF=p

^r

-^F— X-

i

*i) Correct perfect Cadence, omitted (or, more properly, suppressed) on
account of the extension to follow.

*2) ** Intercepted " form of the perfect Cadence (comp. Ex. 32, note *4)).
It is expanded to two measures.

*3) This, and the seven measures which follow, constitute a "new Cadence-
member" (see par. 2gd; Ex. 42).

See also: Mendklssohk, "Songs without Words *% No. 23, meas. 1-9
from the first double-bar; compare carefully with meas. 25-10 from the end.
Also No. 27, meas. 16-4 from the end (cited in Ex. 32-2); No. 31, meas. i-ii
(4-measure Period, repeated, and extended at its end) ; No. 38, meas. i-io.

Chopin, Pfte. Concerto, e minor (op. 11), Romance j meas. 13-22 (quaint
repetition of 2nd half of Consequent Phrase).

Par. 48. THE EXTENSIONS AT THE END OF EITHER PHRASE.

Si

48. An extension at the end of the Antecedent Phrase will
be, as intimated, a comparatively rare occurrence, owing to the
danger it involves of severing the continuity of the two Phrases.

It can scarcely be any more than a brief repetition, or expan-
sion, of the chords which constitute the Semicadence (which it
must confirm, and not destroy), or of the entire semicadence-mem-
ber ; and it will generally be balanced by a similar manipulation of
the final cadence (of the Consequent) also. For example :

AUegretto. Cad

I

I

te

:i. if

iM rrpj JTD

I

r

?

r

^

r

m

mp dolee.

^

J

^

3=5

A

Ab V

.extension.*!)

Consequent.

Bkahms.

*i) Or, more accurately (in both cases), an "expansion" of the Cadence-
tone in the melody.

*2) This Cadence is incomplete, because the new section which follows is
to be entered without interruption. (See the Note following par. 93a.) This
example, from the third movement of the first Symphony of Brahms, is a
unique illustration of ** opposite construction'* (39^).

82

EXTENSIONS IN THE COURSE.

Par. 49.

See also: Brahms, op. 117, No. 3, meas. i-io, 11-20; same work, second
tempo (/li rnoto ed espress.\ meas. i-io, 11-20, 21-30. These Periods are all
of parallel construction, and all extended in the same way, — at the Cadences.
Also op. 118, No. I, meas. i-io.

See also : Mendelssohn, " Songs without Words " No. 37, meas. 30-39
(extension at end of Antecedent only).

Schubert, ** Winterreise '*, No. 14, last fifteen measures (extension at
end of each Phrase); "Winterreise", No. 19, meas. 6-13 (two 3-measure
Phrases; followed by modified repetition, meas. 14-21); ** Winterreise", No.
22, meas. 5-18 (s-measure Phrases), and meas. 16-5 from the end.

Chopin, Mazurka No. 20, meas. 9-24 (repeated Consequent).

6. Extensions in the Course.

49. This mode of enlargement is applied most frequently,
and, at all events, most extensively, to the Consequent Phrase (see
par. 30). The details are enumerated in paragraph 29, which re-
view. For illustration :

Adagio,

ss«

^^

i:

B

r

-f^ f ,-g:

*i)

'^^

*2)

^m

Conseqaent.

Haydn.

Par. 50.

EXTENSIONS IN THE COURSE.

83

*i) The Semicadence is not bridged over, in this case, for the obvious pur-
pose of rendering the following extension perfectly distinct.

*2) Modified repetition of second half of the Antecedent Phrase.

*3) A somewhat disguised sequence of the first half of the Consequent
Phrase.

See also: Mendelssohn, ** Songs without Words *', No. 11, meas. 22%-^
from the end ; a 4-measure Antecedent, and a Consequent containing course-
extensions (in its first and third measures) ; the perfect Cadence is twice evaded
(in measures six and eight of Consequent Phrase), and, finally, a "new cadence-
member" of two measures closes the Period.

50. Such extensions in the course of either Phrase are not
unlikely, here again, to result in '* chain-phrase " formation, as
explained in paragraph 31. As usual, this is far most likely to be
the case with the Consequent Phrase.

See Mendelssohn, ** Songs without Words '% No. 10, meas. n-24 (a
4-measure Antecedent) ; and No. 38, meas. 22)^-41 (ending with an imperfect
Cadence, before final Phrase of the Song).

Chopin, Prelude, op. 28, No. i, meas. 1-25.

But this may even occur in the Antecedent, as the following
(rare) example shows :

Allegro.

I

m

&

1 ^

-0 H \

m

^

^

m

56. {

«

^ -^ 1 dhf^ f^^

i±

:p=f

1

P

m^

tit

J \i

^

*t^

^^^

1

^

84

THE CODETTA,

Par. 51.

j ^.i^'ffL'irVtjifjgTa' i

V ^ ^^

J

Consequent.

I

Bebthovbn. *2)

fr"j. J.

«

3

10

^

ga-r

J.

rr"

^H

etc.

3212:

f I'' I, i -r ii

t

*i) The Antecedent, ending with this emphatic semicadence, is thirteen
measures long. Its thematic evolution is indicated by the lettered brackets.
The Consequent which follows is an example of the 5-measure Phrase.

♦2) Pfte. Sonata, op. 49, No. i, Finale^ meas. 293^-12 from the end.

7. The Codetta.

5 1 . To the class of extensions at the end belongs, further-
more, the Codetta (i. e., little Coda) ; but it differs from the
extensions explained in paragraphs 24, 25, 26, and 47, in being
more independent of the Phrase or Period to which it is added.
For in the majority of cases it follows after the complete perfect
Cadence has been made^ and becoines by that means detached from
its Phrase or Period. (It is most nearly, though not strictly,
analogous to paragraphs 26^, c, d).

Par. 61.

THE CODETTA,

85

For this reason it need not be thematically related to the
Melody of the Cadence-member, but may derive its Melody from
any member of its Phrase or Period ; or it may even consist of an
entirely new, though strictly kindred (organically related), melodic
motive. Review also paragraph 32.

Its harmonic basis may be a reiteration, in almost any metric
form, of the Dominant and Tonic chords of the Cadence ;

Or it may be an optional line of chords, upon the Tonic organ-
point ;

Or it may incline towards the Subdominant chords and keys,
as broader exposition of the Plagal ending (26d). The latter is
probably the most common.

The length of the Codetta, as applied to any of the foregoing
forms, while optional, should not as a rule exceed two measures ;
but it may be, and usually is, repeated, and even extended.

It is very rarely added to anything smaller than the Period-
form.

For illustration, to the Period given in Example 50, the fol-
lowing Codetta is added :

Cad.

87. <

Codetta.

Rep.

Bbbthovrv.

*i) This is the Cadence-measure, bridged over as shown in paragraph 18.

*2) The following two measures cannot be called an ** extension of the
foregoing Phrase, or Cadence-member", for they are melodically independent
of the latter. Hence the independent term Codetta (appendix), to indicate
the independent addition.

♦3) In the original, (Beethoven, Pfte. Sonata, op. 10, No. i, end of
Adagio^ six more measures follow, as additional repetition and extension of
this Codetta ; but they are necessitated simply by the proportions of the entire
Movement^ and are therefore omitted here.

Furthermore, to Ex. 41, No. 3, is added the following:

86

THE CODETTA.

Par.SS.

Cad.

Codetta.

.

i

^

r

£

<Ez.41. No.3.)|

^

lI^

Rep.

3

M:

^

'l^iT

»

^

ffi=S

dim.

i

<

(1

fc

J.'

r - r •
J.

S

^

r

gr^

^'^ Mendelssohn.

1:12

-?-T^

LT

ritord.

n*^

i

=^"

*1)

/T\

f"

f=^

i

*i) The form " tapers " to the end, from the large 2-measure Phrase
which precedes, — through the i-measure Codetta, its repetition, and the
half-measure extension, down to the single Tonic chord, which fades away as
its resonance diminishes, without regard to measure. And the ** ritard.^^
adds emphasis to the purpose of gradual relaxation.

See alto: Mendelssohn, ** Songs without Words '^ No. i, last six meas-
ures (Codetta to preceding seven measures, with which it is to be compared);
No. 13, last four measures (with preceding eight measures); flo. 20, last five
measures (with preceding twelve measures: Phrase repeated and extended);
No. 27, last four measures; No. 40, last six measures.

Chopin, Prelude, op. 28, No. 21, last fourteen measures (two Codettas).
And Ex. 70, note *4).

52« The Codetta is the counterpart of the Introduction or

Introductory Phrase ; and, just as the latter may deteriorate, by

separation from its Period, into a " Prelude ", so the Codetta may

become a

''Postlude''

by still more complete separation from the sentence which pre-
cedes it, and by greater difference of style. Compare paragraph 45.
The Postlude is more common in larger forms, than in the
Period. It is most likely to occur when the form began with a
Prelude, and will, in that case, usually correspond exactly to the
laitter.

See Mendelssohn, ** Songs without Words", Nos. 4, 9, 16, 23, 28, 35,
41, beginning and ending of each.

Par.SS. THE CODETTA. 87

53- From all the foregoing it is evident that the processes of
Phrase- or Period-extension are employed preponderantly in the
second half, i. e., in the later course^ towards, or at, the end of the
sentence.

The first obligation of the composer is to state the leading
musical thought clearly^ simply^ without confusing modifications.
The variation, elaboration, and, most especially, the development
and enlargement of this leading thought will follow, earlier or
later, in fulfilment of the conditions of variety in unity. Review,
briefly, paragraphs 30, 42, 47.

A peculiarly significant reason for extensions at the very end is
touched upon in paragraph 27, and Example 58, note *i).

MISCELLANEOUS EXAMPLES OF PERIOD-EXTENSION.

Chopin. Pr61ude, op. 28, No. i (Antecedent Phrase, eight measures;
Consequent Phrase, chain-form, sixteen measures; Codetta I, two measures
and repeated; Codetta II, one measure, repeated three times; expanded Tonic
at end).

Schubert. "Die schone MUllerin'% No. 8, entire (4-measure Prelude ;
two 6-measure Phrases, Consequent extended ; 2-measure Codetta).

Beethoven. Pfte. Sonata, op. 10, No. i, Finale^ meas. 17-28.

Beethoven, Pfte. Sonata, op. 81, Finale^ meas. 37-52 (repetition of whole
Period); also the following measures, 53-81, up to the double-bar (repetition,
extension, and Codetta).

Beethoven. Pfte. Sonata, op. 31, No. 3, first movement, meas. 46-68
(repetition with intermediate Interlude ; Codetta).

Brahms. Op. 116, No. i (Capriccio), meas. 1-19 (repeated and extended),

Brahms. Op. 116, No. 2 (Intermezzo), meas. 1-18 (extended and re-
peated, with instructive modifications ; see also the last twenty-one measures).

Brahms. Op. 116, No. 5, First Part (extension at end).

Chopin. Mazurka No. 3, meas. 49-64 (quaint repetition).

Chopin. Mazurka No. 17, meas. 1-20 (Introduction and repetition).

Chopin. Mazurka No. 20, meas. 1-24 (Introduction and extensions); also
meas. 41-56 (repetition).

EXERCISB 12.

Former Periods, or new ones, with Introduction or Introductory Phrase
(par. 44, 45, 46).

The same, with extensions at end of either Phrase, or of both Phrases (par.
47, 48).

Here, again, it will be wise to write out the Period first in its primary
(unextended) form, and to elaborate it during its repetition.

88

THE PERIOD WITH CON SEQJLJE NT-GROUP.

Par. 56.

BXBRCISB 13.

The same, with extensions in the course of either Phrase or both Phrases
(par. 49, 50).

The same, with Codetta (par. 51, 52).

These extensions may, finally, all be introduced into one and the same
elaborate Period (illustrating the entire material of paragraphs 44, 47, 49, 51^
in one example).

Write out each Period first in its unextended form.

CHAPTER VII.

GROUP-FORMATIONS.
The Period with Conseqjljent-Group.

54. The term *' Consequent-group " has been adopted by
the author as a substitute for "Consequent-repetition", in appli-
cation to those forms in which the reproduction of the Consequent
phrase is modified to such an extent, or in such a manner, that the
epithet "Repetition" is not strictly permissible. It is to all
intents and purposes a reproduction of the second Phrase of the
Period, but it is not a simple (or even unessentially modified)
repetition. The term indicates that the two phrases thus obtained
constitute a double version of such unquestionably uniform con-
tents as to be virtually one and the same Consequent (a " Double-
Consequent ").

55. The details of the apparently subtle distinction between
such a "reproduction" and a genuine "repetition", are as
follows :

ist, any essential difference in the harmonic formation of the
two Cadences (of the two versions of the Consequent Phrase) would
antagonize the principle of " repetition ", as defined in par. 21
(which carefully review) . For illustration :

Andante espressivo.

69.

5^

Antecedent Phrase.

For Accompaniment, see Original.)

Par. 56.

THE PERIOD WITH CONSEQUENT-GROUP.

89

Semi-cad.

Ist version of

I

m

r^^^i

a

-^-r

=^

Conseqiieiit Phrase.

te

^

m^=^

3.

-9 — ^

, 5

frt

ff-Tr

J J

s

^

i — r

I

I

L.

.L=i

*i>

# »-

TT-r

^t-^-m.

^

^

.^ 1 •

Y

2nd version of Conse-

iJ.

Frf

«^

T — rr

-0 — #-

f

-?*^

^

r^

¥

ifc

qnent Phrase.

{

fe

I I ! ! ~ I IN I .. I I

t ^ J j hi J - kJ— J^ -aJ a*^

'M

-.-U

^

^

?=^

4

^^

*-M^

s

■r T"

'-B

f

Mendelssohn. No. 25.

I3 expanded

(Gad.-ezpansion.)

90

THE PERIOD WITH CONSEQUENT-GROUP.

Par. 56.

♦i) This first Cadence of the Consequent Phrase is made upon the Relative
Tonic, (e minor, instead of G major,) and thus the idea of ** repetition " is
already frustrated, because the final Cadence is to be made upon the original
Tonic, G major.

♦2) The similarity between the two versions of the Consequent Phrase is
so great that it appears at first glance to be simply Phrase-repetition. But
closer scrutiny of the two essentially different Cadences reveals the necessity
for the above distinction. It is not a repeated Consequent, but a ** Group of
Consequents." The importance of this distinction, as defined by the forma-
tion of two successive cadences, is also shown in paragraph 39a, where it is
the obligatory condition which distinguishes the ** parallel Period " from the
"repeated Phrase", or, in a word, the ** repetition " from the "reproduc-
tion ". See also Ex. 70, note *2).

See also: Mendelssohn, ** Songs without Words", No. 14, meas. 11-23
from the end (final Cadence modified by chord-Fifth in Bass, on account of
thematic connection with following Phrase) .

Mozart, Pfte. Sonata, No. 8 (Cotta ed.). Andante^ meas. 28-43 (second
version of Consequent extended at end).

Compare these examples carefully with the illustrations given in Ex. 50 ;
Ex. 51; Ex. 62, note *2) ; with especial reference to the condition of the
Cadences. The latter are examples of Consequent and Antecedent repetition^
because the cadences are essentially, if not exactly, the same ; and are not to
be confounded with the above (Ex. 59), which is a Consequent-^f^«/, because
of the essential difference of cadences.

2ndly, a reproduction of the Consequent Phrase upon other
scale-steps^ as sequence, should not be called a repetition, even if
the melodic design remain exactly, or nearly, the same. A
''sequence" is a "reproduction", but never a "repetition" in
the strict sense here necessarily maintained. For example :

60.

%> Y

t=t

Antecedent Phrase.

(For Accompaniment, see Orif^nal) *1)

i

W

W

I

1st version of Consequent Phrase.

i?c

5CZ?:

^

^— #-

i

"0,

t=t

Semi -cad.

P'h'i> n.'a.i

Par. 65.

THE PERIOD WITH CONSEQJ7ENT-GROUP.

91

2nd version of Consequent Phrase*

Bbbthovbn.

*i) Pfte. Sonata, op. 2, No. i, Finale^ first twelve measures.

*2) The third of these three Phrases is mainly a sequence of the second
one, and with a "reversed" Cadence-member; consequently, the form is
** Period with Consequent-group *'.

See Chopin, Pr61ude, op. 28, No. 24, first nineteen measures (Introduc-
tion two measures ; Antecedent-Phrase four measures, wLth modified repeti-
tion ; Consequent-Phrase first version, five measures ; second version, partial
sequence, four measures).

3rdly, the term Consequent-group should be substituted for
Consequent- re petition in those cases where the reproduction em-
braces more elaborate modifications (of what is, nevertheless, unmis-
takably the same Phrase-contents) than would be supposable or
permissible in an ordinary *' repetition " ; compare, carefully, all
of paragraph 19. For illustration :

61.

Lento*

^^

^S

Antecedent Phra«e.

*

t=t

*

OlJ

p

4

>j»fc i ^trrr' fr wj? injTf;7 1 :

Ist version of Consequent.

m

= wi~^^

t^tp

^-^ *

lyu

92

THE PERIOD WITH ANTKCKDKNT-GROUP.

Par. 56.

Chopin.

^^E

2nd version of Consequent.
♦1)

^

^ ft"^

•2)

W f

t

^

*i) There can be no doubt of the identity of this Phrase as second version
of the preceding- Consequent Phrase^ although it diverges from the latter more
and more, in its course, and ends with a totally different Cadence.

*2) See also the twelve measures which follow, in the Original (Chopin,
Mazurka No. ii, op. 17, No. 2).

And Mendelsohn, "Songs without Words", No. i, last twenty-one
measures (Antecedent Phrase four measures ; first Consequent four measures ;
second Consequent, eight measures; Codetta, five measures) . — No. 31, meas.
143^-5 from the end (Antecedent, two measures; first Consequent, two
measures ; second Consequent, partly sequential, largely expanded, six meas-
ures). — No. 33, last twenty-three-and-one-half measures (Antecedent, four
measures ; first Consequent, four measures ; second Consequent, ten measures,
analyzed in Ex. 42, No. 2; Codetta, five measures).

Schubert, ** Winterreise'*, No. 15, meas. 19-6 from the end.

56. By exactly the same processes, the Antecedent Phrase of
a Period may also be reproduced in this group-form. But the
'* Antecedent-group " is far less common than the Consequent-
group. See, again, paragraphs 53, 43. One illustration will
suffice :

A lUgretto.

Ist version of Antecedent.

62.

r^

-w—^

g

y 2nd

IS'-T-

^

■i9-

'D

I

^^

^^

(

version of Antecedent.

II ^ n

m

i i\i i i

V

^^

9^=fE!

A.

A

1

g

Par. 67. ^i

THE PERIOD WITH ANTECEDENT-GROUP,

93

Conseqaent.

i

Bepetition of Consequent.

m

f^

J i j i"] J i .nj

SuABiAN Folk-song.

f

^

f

^

-T) j J . J -I -^ .^ -* .J

i

1^

r

^^

ir

♦i) The second version is a ** sequence'* of the first, and, consequently,
ftot a "repetition". It is an Antecedent-^^«/. Compare 55, 2ndly.

*2) Here, on the contrary, a genuine repetition of the Consequent takes
place, because (besides entire coincidence of melody) the Cadence in measure
thirteen is only a concealed form of that in the final measure, without essen-
tial change.

*3) This measure is a quaint and characteristic expansion of the (strictly
speaking, unaccented) first beat, — a poising of the voice upon the ** up-beat"
before the rhythmic movement starts into life. This accounts for the irregu-
lar (5-measure) formation.

See also: Beethoven, Pfte. Sonata, op. 2, No. i, MenuettOy first fourteen
measures (first Antecedent- Phrase, four measures; second Antecedent,
sequence, four measures extended to six; Consequent, only two measures, but
repeated).

Haydn, Pfte. Sonata, No. 14 (Cotta ed.), first twelve measures.

Haydn, Symphony in B|> (Peters ed.. No. 12) Finale, first twelve
measures.

Chopin, Mazurka, No. 36, first twelve measures.

Chopin, Pr6lude, op. 28, No. 5, first sixteen measures (Introduction, four
measures ; first Antecedent, four measures ; second Antecedent, sequence, four
measures; Consequent, four measures) — ; the remaining twenty-three meas-
ures of the same Pr61ude are similarly constructed (the last seven measures
are Codetta).

Brahms, op. 119, No. 3, l^st thirty measures (first Antecedent, four meas-
ures ; second Antecedent, four measures ; Consequent, a long Chain-Phrase to
end).

The Phrase-Group.

5T. The Phrase-group is a series of Phrases (at least three
in number), in coherent succession, and of similar or kindred

94

THE PHRASE-GROUP.

Par. 57.

character and style ; but either too similar to, or else too indc'
pendent of, each other, to exhibit the distinctive condition of the
Period-form, viz., the opposition of a mutually dependent thesis
and antithesis (see par. 37, last clause).

The Group will never contain less than three Phrases, because
two such, if coherent (not separated, as in Ex. 74), will invariably
represent the ./^^r^W- relation, in one or another of its many possi-
ble phases. On the other hand, the Group may embrace as many
more than three Phrases as can be added without sacrificing
coherency, or extending to an absurd extreme of ** formlessness ".

Each of the Phrases should close with a light semicadence (com-
pare pars. 36, and ^yi)* A perfect Tonic cadence (in any key) of
sufficient cadential force to complete its Phrase, can occur only at
the end of the Group. There are some rare exceptions to this
strict rule, but they do not concern the beginner. (This is in
accurate keeping with the rules given in paragraph 31 for the
Chain-Phrase, which review. The '* Group of Phrdses " is simply
a larger growth of that * ' Group of Members ' ' which constitutes
the Chain-Phrase).

The following example illustrates the Group of similar
Phrases :

63.

A ndanie tranq.

Phrase No. 1.

riji i rsg

« — I — m — m • — « — #

etc. (See Original.)

g

^^^

s

Phrase No. 2.

y^^J ^ \ ^, ^ \ ^iU^^ .

*1)

Phrase
No. 8.

Par. 58.

THE PHRASE-GROUP.

95

Mendelssohn. No. 33.

' juii ' tii ^

«3) *4)

^P

*i) The similarity between these three Phrases is strikingly exhibited in
the first half of each, — the melodic members being almost identical.

*2) The Bass makes this irregular leap upward from the Leading-tone (a)>
simply in order to evade the complete Tonic cadence.

*3) This final Cadence is unusually brief, it is true; but it, nevertheless^
conforms to the conditions of the perfect Cadence fully enough to check the
harmonic and melodic current, and bring the ** Group " to an end. (Par. 93a.)

*4) Comparison of this example with the foregoing, reveals the charac-
teristic difference between the ''''Period with Consequent- or Antecedent-
Group ", and the single ** Phrase-Group *'. Ex. 59, for instance, also consists of
three Phrases (like the above) ; but the number is reducible to the original two
of the Period-relation^ because the second and third are no more, in sub-
stance, than one (Consequent) Phrase, together. In Exs. 60 and 61 the same
reasoning applies; and in Ex.62 all four Phrases are similarly reducible to
the original two. But in the above example (63), on the contrary, such a
reduction to two Phrases is not tenable, because the entire sentence is simply
three different versions of one primary phrase.

58. The Group of independent (or dissimilar) Phrases, on
the other hand, is probably more common. The reduction of the
series to two primary Phrases will not be possible, on account of
the thematic individuality of the several members of the Group ;
but at least general resemblance^ and close organic relation, must
be preserved ; and, as before, no complete perfect cadence should
occur until the end is reached. For example :

64.

Adagio.

^

4

^- f ^-^

Phrase No. 1.

mf

^

3n. t

*r^^--=

"I

* I II f ^ > II i

•1)

^ -...., >o-i #^r^ #

I . t g >

i

*l} It i< erident that this series, again, cannot be reduced to tsro Phrases
and be thus demonstrated as an extended " Period " of some kind, for each of
the three Phrases is an independent melodic factor of the collectiTe sentence,
ttumt^ perfect organic cohesion is maintained (chieflj through the uniform
flccfympani ment) ,

H^.e also the following general illnstrations of the Phrase-group (both sim-
ilar and dissimilar) : Meki>elssohx, "Songs without Word*? '. No. 13, last
twenty-three measures (four Phrases, Nos. i and 2 similar — almost repetition ;
No, 4 extended four measures at end; Codetta). — No. 16, nieas. 4-9 (three
2'meik%ure Phrases). — No. 2o,meas. 28-17 from the end, — ^o. 26, meas. 28>^'iz

/.

Par. 00. THE ELISION. 97

from the end (Phrase i, two measures; Phrase 3, two measures; Phrase 3, four
measures; Phrase 4» two measures; Phrase 5, two measures; Phrases 4 and 5
repeated and extended). — No. 41, meas. 15-5 from the end (three Phrases; No.
3, three measures). — No. 3, meas. 5-29 (six Phrases).

Chopin, Mazurka, No. 33, first twenty-two measures (five Phrases, /ar//y
similar, the last one extended).
x^ Chopin, Prelude, op. 28, No. 9 (three similar and regular 4-measure
Phrases). Prelude, op. 28, No. 2 (three Phrases, similar, each expanded).

Schubert, Pfte. Sonata, No. 4 (op. 122, Menuetto^ first twelve measures.
Pfte. Sonata, No. 8 (C minor). First movement, meas. 40-53 (three similar
Phrases, of unequal length ; entire Group repeated, in the following fourteen
measures) .

Grieo, Lyric Pieces, op. 12, No. i (three Phrases; complete repetition).

Brahms, op. 118, No. 6, first twenty measures (five similar Phrases, No. 2
a repetition of No. i) ; the following twenty measures are a modified repeti-
tion of this Group; the following eighteen measures are a Group of four
Phrases (i and 2 regular, 3 extended, 4 a sequence of 3) ; the following
measures, up to the end, are a modified reproduction of the first Group.

To what length the Group of Phrases may possibly extend, may be seen
by reference to paragraph 104. But the pupil must carefully shun such
difficult experiments for a time. And he must look upon the idea of the
Group-form with distrust, in any case, as being the type, — not of good con-
sistent ** Form", — but rather of " Formlessness ".

59. This same idea is also applied (though far more rarely) to the
Period^ of which similar Groups may be formed, especially when the Periods
are small (four measures) . Such Period-groups will present the appearance,
usually, of large Phrases with a slight semicadence in their center; and it is
possible that this break may be present in some and absent in other of the
Members.

See Schubert, Pfte. Sonata, No. 7 (op. 164), Finale^ first thirty measures
(each Period expanded ; the second one extended at end).

Mendelssohn, *' Songs without Words ", No. 32, meas. 16-28.

Haydn, Symphony 5 (Peters ed.)» Third movement, Trio^ first twenty-six
measures.

The Elision.

60« When the Cadence of a Phrase or Period corresponds
harmonically so accurately to the beginning of the following Phrase
or Period that end and beginning are practically identical, it is
possible to suppress the entire Cadence-measure^ and hasten on into
the next Phrase without the pause or check for which Cadences are
expressly made (par. 2^).

This suppression or ' ' Elision ' ' of the Cadence is almost
exclusively applied to the perfect Cadence, though possible at semi-

98

THE BLISION.

Par. 60.

cadences also. It may be effected, firstly, when the approach to the
Cadence has been so indicative or suggestive of the latter that its
absence will be sufficiently compensated for ;

Or, secondly, when the beginning of the succeeding Phrase is
striking enough to remove any misconception of the form (see par.
93a, first and last clause).

The object of the elision is either to avoid an unnecessary and stagnating
pause, or to obtain an exhilarating, urging effect, which is sometimes very
effective. But it must be distinctly understood that it is a comparatively rare
and dangerous artifice, the expediency of which should be carefully tested in
each case. For illustration t

6S.

I. Andante,

Haydn.

i

k

a. Allegretto.

4 4 ^r~4 \

(See Original.) *3)

«

-# #■

^

df

-0-0-

^

*rt

m^

i

Par. 60.

THE ELISION.

99

JKepetition(extended.)

Mendelssohn.

*i) These small notes represent the perfect Cadence which is due and ex^
pected at this point, but which is suppressed (omitted), causing the elision of
the entire Cadence-measure. In other words, this is the fourth measure of the
first phrase, and, simultaneously, the first measure of the following one. The
latter immediately asserts itself so completely (through abrupt change of style
and abrupt transition from/ to y) that the actual cadence is "cut out"
entirely, and the sum of measures in the two Phrases is only seven, instead of
eight.

*2) The first chord in the fourth measure represents the " evaded '* form
of the (expected) cadence; but, at the same time, it corresponds exactly to
the beginning of the first melodic member of the Phrase, which it is the
author's intention to repeat, and which he proceeds to repeat immediately^
without the — in this instance undesirable — delay which a full exposition of the
cadence would cause.

*3) "Songs without Words,*' No. 42, near the end; the "extension*' is
cited in Ex. 41, No. 2. See also Ex. 70, Note *4).

Beethoven, Pfte. Son. op. 2, No. 3, Adagio, between meas. 10 and 11.

Mendelssohn, " Songs without Words " No. 4, bar 5 from the end. — No.
9, bar 3 from the end. — No. 11, between meas. 4 and 5 (compare with measure 19
from the end, where the Elision does not appear). — No. 31, bar 5 from the end.

Mozart, Pfte. Son. No. 11 (Cotta edition), between measures 7 and 8;
also last movement {Rondo), measures 13-19 (repeated Phrase with Elision).

On the other hand, care must be taken not to confound the
Elision with those forms of concealed or evaded cadence, where the
same abrupt leap into the new style of the foUowng Phrase takes
place within the Cadence-measure^ but, contrary to appearances,
does not suppress the latter. For example :

m

Allegro.

66. {

I

g^-~H

r-X

^F-*

S=l2f

w

m

^^s

3

^

^§

i

lOO

THE ELISION.

Par. 60.

f tr r . a

Mozart.

*i) Notwithstanding the abrupt and misleading change of character at the
beginning of this cadence-measure, there is no Elision, for the coming Phrase
contains its four measures without this one. The " change" must be regarded
as occurring upon the second i6th-note, not upon the first one ; consequently,
according to par. 2a (third clause, which review), this is only the preliminary
measure of the 2nd Phrase; and, though thematically consistent with what
follows, it is no more than one of the innumerable modes of ** bridging over"
the Cadence-space (par. i8).

Beethoven, Pfte. Son. op. 2, No. 3, meas. 12-21, is a still more delusive
illustration of the absence of an Elision where one is apt to be suspected
(between measures 12 and 13). But, here again, the new Phrase (and its subse-
quent reproduction, also) begins upon the second i6th-note of the cadence-
measure, and has its quota of 4 measures without the ** bridging", though the
latter is directly characteristic of the coming Phrase. Of all the possible ways
of bridging over the space between one Phrase and the next, this must be con-
ceded to be the most consistent and admirable ; i. e., with material which antici-
pates the coming Phrase.

See also Mendelssohn, ** Songs without Words", No. 28, meas. 5 from the
end; here, also, "end" and ** beginning" are interlocked, but there is no
Elision. — No. 26, meas. 11 from the end; no Elision,— -comp. 2 meas. later.

EXERCISE 14.

Former Periods, or new ones, with Consequent-group, in each of the three
varieties of structure explained in par. 55.

The same, with Antecedent-group (par. 56).

Work from Cadence to Cadence ; make all Cadences sufficiently emphaticy
— rather too distinct than too vague. Still avoid, for a time, any elaborate
modes of evading or concealing the Cadences, such as may too completely dis-
guise their purpose, and impair firm outlines of form. — Review the directions
ffiven in Exercise 9,

:^

Par. 62. THE DOUBLE PERIOD. lOI

EXERCISE IS.

Write a number of Phrase-groups, of both similar and dissimilar thematic
contents, according to par. 57 and 58. Each group must be limited to tAree
fundamental Phrases, at present, though this number may be optionally in-
creased by the repetition (exact or modified) of any of the Phrases.

And make a few experiments with the Elision ; best in connection with
repetition of Phrase or Period.

CHAPTER VIII.

THE DOUBLE PERIOD.

6 1 . The Double period consists in the union of two periods,
and embraces, consequently (when regular), four Phrases^ so con-
ceived and distributed that the Period-relation is apparent between
Phrases i and 2, between Phrases 3 and 4, and also, on a broader
scale, between these two pairs.

62. The two Periods of a legitimate Double'iperiod, form are
just as coherent, and just as closely dependent one upon the other,
as the Antecedent and Consequent Phrases of the simple Period
forms are. And, for this reason, the Cadence in the center (i. e., at
the end of the 2nd Phrase) must be in the nature of a Semicadence^
though almost unavoidably somewhat heavier than an ordinary light

semicadence.

The Double period must be conceived simply as an expanded growth of
the single Period, the Phrases of which assume such a breadth (length) that
they almost necessarily separate slightly in their respective centers. These in-
termediate points of repose will, naturally, be lighter than the one in the
center of the whole. The latter, though still only a Semicadence (inasmuch as
it corresponds to the semicadence of the simple Period), will be rhythmically,
and perhaps even harmonically, stronger or heavier than those at the termina-
tion of the I St and 3rd Phrases.

The various degrees of rhythmic and harmonic weight which it is possible
to impart to a semicadence-chord, are distinctions which increase in import-
ance as the forms enlarge. Therefore, the following general rules must be
observed : A semicadence is harmonically ** heaviest" when made upon some
Dominant Triads or upon the Tonic Triad of a related key, as these are strong
chords, but chords which, while they invite repose, still suggest inevitable
onward movement. A semicadence upon any form of any other chord, or upon
the dissonant forms of Dominant (or other) harmony, is *Mighter'% either
because it presses forward too vigorously of itself, or because, if itself a heavy,

I02 THB DOUBLE PERIOD. Par.61

Stagnating chord, it gives immediate birth to an active counteracting agency.
Further : a semicadence is rhythmically heavy or light in exact proportion to
the duration of the chord, — the length of time spent in pausing, — and the con-
sequent length or brevity of the "bridging" which follows. Finally: the
rhythmic distinction is far the more important of the two.

63. The Cadence-conditions of the regular Double period are,

then, as follows :

(a) At the end of the first Phrase a light semicadence, con-
sisting perhaps most frequently of the Tonic chord, with Third or
Fifth (instead of Root) in one of the outer parts ; or the Domi-
nant chord, consonant or dissonant, — ^the latter being lighter than
the former ; or possibly some other chord in one of its lighter
forms. And, in any case, rhythmically more or less brief, exactly
according to the degree of interruption desired (see par. 2^, final
clause) .

(b) At the end of the 2nd Phrase a heavier semicadence,
corresponding generally, in choice of chord, to the table given in
par. 36 (to which strict attention must be given), but probably with
more rhythmic stress, i. e., longer duration, than in the single
Period. It is most apt to be a Tonic chord of the Dominant Key;
or, more rarely, a Tonic chord of the Relative, or some other next-
related key, approached with a more elaborate and emphatic modu-
lation. But it must never be so heavy as to lose its semicadence
effect^ and thus destroy the continuity of the whole.

(c) At the end of the 3rd Phrase a light semicadence again,
perhaps corresponding to the first one, though a little more likely to
incline toward the Suhdominant harmonies ; and again, with a more
emphatic modulatory manifestation.

(d) At the end of the 4th Phrase (in the regular, complete
Double period) the Tonic perfect Cadence.

64. The necessary cohesion between the two Periods is most
effectively preserved by constructing the 2nd Period melodically
PARALLEL with the I St. See par. 39a.

The parallelism need not, however, be maintained any farther
than during the first melodic member of the 2nd Period ; although,
generally, the entire 3rd Phrase corresponds exactly or nearly to the

Par. 64.

THE DOUBLE PERIOD.

103

I St Phrase ; and, in cases of extreme parallelism, even the 4th
Phrase follows the thematic design of the 2nd Phrase, nearly to the
end.

The following example illustrates the Double period as an " ex-
pansion of the single Period " , in parallel construction :

Phrase 2.

Andante,

Period 1. PhraM 1.

67. <

V Period 2.
Phrase 3 (like 1).

! I 1 r^

I s s »s s -

^. J J.3 J."^ I

T

^^-^TtT'^^-^

Beethoven.

^

JtJ?^**.

LU-L^

m

^m

F | 7 *r f

^

♦i) This cadence, like that which precedes it in the 4th measure, is. made
upon the Dominant harmony with the chord-fifth in the Soprano. But it is
rhythmically stronger here than in the 4th measure.

*2) The parallelism of construction extends up to this point, with no other
than the purely unessential modification in the 9th measure.

*3) There is no reasonable doubt of this being the Z>ouble'penod form ;
but, the measures being small and the tempo somewhat active, it creates the
impression of a Large single Period (of 8-measure Phrases and parallel con-
struction) with a semicadence in the middle of each half, coinciding, on a
larger scale, with the syntax of Ex. 13, No. 3 — ^which see. This latter analysis
(i. e.. Large single Period) would certainly be correct if the tempo were
allegro or presto.

I04

THB DOUBLE PERIOD.

Par. (IB.

65. The transitional grades and progressive shades of distinction from
the perfectly unmistakable Single-period form with no intermediate interrup-
tions, into the perfectly unmistakable Double period with its four legitimate
cadences, are so innumerable, that it will often be impossible to define accu-
rately the denomination of certain examples of this kind found in musical
literature ; and even so slight a thing as a deviation in the tempo or ** phrasing'*
of the performer may influence, perhaps positively alter, the structural im-^
pression and analysis. Of one thing the student may rest assured, namely: in
proportion as the distinction becomes thus more minute and questionable, it
becomes of less and less moment to define it, and less 'wise to insist upon an
absolute definition. See par. lo; and par. 77.

For examples of doubtful form, between Single and Double period, see :

Mendelssohn, ** Song without Words '* No. 12, meas. 7-22 (a noticeable
semicadence in the center of the first half, but none 'whatever in the second half;
like Ex. 13, No. i). No. 22, first 9 measures (exactly the same). No. 8, first 17
measures (probably Double, despite the rapid tempo). No. 19, measures 3- 11
(probably Double). No. 25, first 10 measures (probably Single; extended at
end). No. 32, measures 4-14 (probably Single; extended at end). No. 48, first
8 measures (probably Single) .

Beethoven, Pfte. Son. op. 10, No. i. Adagio, first 16 measures (probably
Double) ; op. 10, No. 3, Menuetto, first 16 measures (probably Single) ; op. 14,
No. I, Allegretto, first 16 measures (probably Single); Rondo, op. 51, No. i,
first 8 measures (probably Single).

Chopin, Pr61udes, op. 28, No. 8, first 8 measures (probably Single).

The following examples, on the contrary, are beyond a doubt
genuine Double periods, of parallel construction :

68. <

Allegro.

mt'us w %r=mw^-

Period 1.
Phrase 1.

Par. 86.

THE DOUBLE PERIOD.

105

.. PRBIOD 2.

1)

K f 'rTf ir"^[r' iLM^g

Phrases (likel).

t^

Bbbthovbn.

*i) This cadence, it is true, cannot be called heavier than the one in the
4th measure; nevertheless, it serves its purpose sufficiently well. The con-
struction of the 2nd Period is here again extremely parallel, almost to the
cadence; but compare with Ex. 49; and observe that the parallel Double
period differs from the repeated Period in the differentiation of its 2nd and
4th cadences.

See also : Beethoven, Pfte. Son. op. 22, Finale^ first 16 measures.

Chopin, Preludes, op. 28, No. 7 ; No. 10 (Phrase 3 sequence of Phrase I ;
Phrase 4 like 2, excepting Cadence; slight cadence-extension) ; No. 14 (Phrase
2 extended to 6 measures ; Phrase 4 only 3 measures ; Codetta 3 measures) ;
No. 23 (Phrase 4 extended to 5 measures ; Codetta, 6 measures, beginning after
— or with — an Elision) ; No. 17, first 18 measures ; No. 19, first 16 measures.

Chopin, Nocturne No. i (op. 9, No. i), first 18 measures (cadence-exten-
sion) ; Nocturne 5 (op. 15, No. 2), first i6 measures (regular, and extremely
parallel, — almost Single period repeated).

Mozart, Pfte. Son. No. 7 (Cotta ed.). Adagio^ first eight measures; Pfte.
Son. No. 13 (Cotta ed.). Finale^ first 16 measures ; Pfte. Son. 14 (Cotta), Finale^
first 16 measures; Pfte. Son. No. 15 (Cotta), 2nd movement, first 16 measures.

Schubert, Pfte. Son. No. io(B|7 major), first 18 measures (the 2nd cadence
expanded); Pfte. Son. No. 2 (op. 53), Scherzo^ first 40 measures (very large;
8-meas. Phrases, Nos. 2 and 4 extended at end to 12 measures).

Mendelssohn, ** Songs without Words '*, No. 15, first 22 measures (Introd.
6 meas.) ; No. 18, first 17 measures ; No. 24, first 18 measures ; No. 30, first
15 measures (4th Phrase short) ; No. 37, first 17 measures.

Brahms, op. 116, No. 6, "Trio'* (5-sharp signature). — Op. 116, No. 7,
first 20 measures (each Consequent extended) ; compare last 31 measures of
same work. — Op. 117, No. i, **/fi Adagio^\ — Op. 117, No. 2, measures 22
(second half) to 38. — Op. 118, No. 5, first 16 measures. — Op. 119, No. r, measures

io6

THE DOUBLE PERIOD,

Par. 66.

1-16. — Op. 119, No. 2, second tempo (4- sharp signature) first 16 measures. —
Op. 119, No. 4, measures 1-20; measures. 2 1-40; measures 41-60 (each a regular
parallel Double period, of 5-measure Phrases).

66. The CONTRASTING construction (39c) of the 2nd Period,
in the Double-period form, is far more unusual and confusing than
the parallel construction ; because, when the 3rd Phrase does not
melodically confirm the ist one, it is much more difficult to preserve
the cohesion of the two Periods (demanded in par. 62), and to pre-
vent the impression of two separate^ independent Periods (as in the
Two-Part Song-form: see par. 74).

Still, **four Phrases" in coherent succession, if such be care-
fully sustained, are likely to be a Double period from sheer symmetry
of design, whether Phrase 3 is like Phrase i or not ; only excepting,
of course, possible cases where the 4 Phrases are reducible^ by clear
evidence of repetition, to a *' repeated Period", or to a ** Group
of 3 Phrases." But the continuity of the Double-period form
must be assured, by avoiding too complete a Cadence at the end of

the 2nd Phrase.

See again Mendelssohn, ** Songs without Words '% No. 13, last 23 meas-
ures, already cited as example of the Group of 4 Phrases.

And Chopin, Prelude op. 28, No. 18 (Phrase i, 4 measures; Phrase 2,
similar, 5 measures ; Phrase 3, four measures ; Phrase 4, extended to 8 meas-
ures). This might be called a Contrasting Double period, but the title ** Phrase-
group" is more consistent with its character.

Furthermore, the Contrasting Double period may represent a
Large contrasting Single period, with interruption in the center of
each half (see par. 62, second clause). For illustration :

Andante.

69. <

&

Period 1.
Phrase 1.

m

p

i-i

£

Par. 67.

THE EXTENSIONS OF THE DOUBLE PERIOD.

107

Mendelssohn.

Phrase 4.

ii^-^-^

^^

I n r i r

^ etc.

t

azv

t

1

*2)

*i) Thematically (i.e., melodically) , Phrase 3 is totally unlike Phrase i;
and there is likewise no resemblance between Phrases 4 and 2. It is '* contrast-
ing'' construction. But oXo^^ formative relation is preserved, throughout, by
uniformity of style and character; and the continuity of structure is absolute.

*2) The "perfect-cadence Tonic" is wanting in Bass, here, because an
extension follows, to which reference will be duly made.

*3) ** Song without Words " No. 2, near the end. Compare these meas-
ures with the first 16 measures of the same Song, where the construction is
parallel.

See also : Mendelssohn, ** Songs without Words ", No. 6, measures 18-34 >
No. 3, first 16 measures after the 2nd double-bar.

Mendelssohn, op. 16, No. i, first 21 measures (4th Phrase short ; Codetta,
2 measures, repeated and extended).

Haydn, Symphony No. 6 (Peters ed.), Menuetto^ first 18 measures (Phrase
4 extended, as shown in Ex. 39, No. i).

Chopin, Mazurka No. 32 (op. 50, No. 3), first 16 measures.

Beethoven, Pfte. Son. op. 22, Adagio^ first 9 measures (a little doubtful
whether Single or Double period, — probably the latter).

The Extensions of the Double period.

67. Nothing further need be said, here, concerning the details
of ordinary Phrase-extension (par. 24 to 30), than that they may,
as a matter of course, be applied to any of the Phrases which collec-
tively constitute the Double period. But see par. 53. The only
devices of enlargement to be specially considered in this place, are
those of modified repetition^ and Group-formation^ which may be
utilized as follows :

Io8 THE EXTENSIONS OF THE DOUBLE PERIOD. Par. 67c.

(a) The modified repetition of the entire Double period, of
which an example will be found in Mozart, Pfte. Son. No. ii
(Cotta ed.), 'Andante^ first 32 measures (rep. variated).

See also : Schubert, Pfte. Son. No. 6 (op. 147), Finale^ measures 51 to 80
(quaintly modified repetition, contracted two measures at end ; but otherwise
preserving all salient traits).

Chopin, Mazurka No. 13, first 36 measures (Introd. 4 measures) ; Mazurka
No. 37, first 44 measures (Phrase 4 each time extended to 10 measures ; repeti-
tion modified) ; Chopin, Nocturne No. 3 (op. 9, No. 3) first 40 measures (Phrase
4 extended to 8 measures ; repetition elaborate) .

(b) The modified repetition of either, or each, of the two
Periods.

This is a somewhat misleading device, especially when applied
to ^<& first Period (comp, par. 42). A Double period, if interrupt-
ed in the center for the sake of repeating its first Period, will
almost certainly present the appearance of two separate Periods,
and become a Two-Part Song-form. But see Chopin, Mazurka
No. 29, first 32 measures (contrasting construction ; each Period
exactly repeated, and evidently without completely disturbing the
continuity of the whole).

See also : Beethoven, Pfte. Son. op. 31, No. 2, Finale^ measures 1-31 (2nd
Period repeated and extended); Pfte. Son. op. 106, Adagio^ measures 1-26
(ditto; Introd. one measure).

(c) The enlargement of the Double period to 5 or more
Phrases by repetition of the last Phrase, or Group-formation of
the final Consequent.

This process of extension can scarcely be applied to any other
than the Double period of parallel construction, in 'which identity
of design is so fully assured by the correspondence of Phrase 3 to
Phrase 1, that it cannot be destroyed by anything that follows (in
the same section).

The repetition of the last Phrase is, of course, imaginable even in a Double
period of contrasting construction. But Gr<?»/-formation would be so diffi-
cult to prove at the end of a contrasting Double period, that the entire sentence
would almost certainly be called a "Phrase-group**, from sheer expediency.

For illustration :

Par. 67c. THE EXTENSIONS OF THE DOUBLE PERIOD.

109

A ndante cantabiU.

70.

I

w

■^-

Pbriod 1.
Phrase 1

s

n

I

J

^n

-8-

-8-

I

I -L I

I I r I

I

^ r- / f r

V

J

P|. ^i^ r r ^ L L ^

J^ J

t

t

(etc. See Original.)

?

-2— #-

.fie-

f.

V

^*

^

i T-'^r^

Phrase 2.

^ in.iij i

^^ ^^ii-1

5xp:

-« ^

V

S

Pbriod 2

Phrased (likel.) «1)

i

•f=- r

^

S^

V

Phrase 4.
(first version).

t^

I I

Phrase 4.
(second version).

i

^

»

i=n.-j

1=1

'iimri

j=m

J

w

-A — 0-

no

THE EXTENSIONS OF THE DOUBLE PERIOD.

Par. 68a.

Codetta.

Chopin. Nocturne, op. 15, No. i.

*i) At this point, already, the identity of the ** Double-period " form is so
fully established, that no reasonable amount of enlargement by the addition of
Phrases (whether similar or not) could change the denomination.

*2) The perfect cadence, due and expected at this place, is completely
evaded. The following Phrase is therefore nol a repetition, but a reproduction,
of the preceding. It is a ** Consequent-group", as shown in par. 54.

*3) Although this is the genuine form of the perfect cadence (Root in both
outer parts), it is made so brief, and followed so abruptly by the new harmony,
that it assumes the nature of the ** concealed cadence " ; the following 2 meas-
ures are a modified repetition of the second half of the Phrase.

*4) At this place there is an unusually distinct ** Elision" of the cadence-
measure. This chord certainly stands for the end of the foregoing Phrase ;
but it is just as certainly, at the same time, the beginning of the Codetta, as
comparison with the first thematic member of the Double period proves.

68a.. The enlargement of the Double period, finally, by the
addition of an Introduction (par. 28), an Introductory Phrase (par.
44), a Prelude (45) — or by the addition of a Codetta or Postlude
(par. 51-52), — is not only possible, but more appropriate and neces-
sary than in smaller forms. Illustrations are contained in some of
the above references, and will also be found in the following

MISCELLANEOUS EXAMPLES OF DOUBLE-PERIOD EXTENSIONS.

Mendelssohn, "Song without Words'' No. 20, first 21 measures (five
4-measure Phrases, No. 5 new, obtained by Group-formation of second Period) ;
No. 21, first 32 measures (Prelude 8 measures; Double period extended by
Group-formation; Phrase 4 three measures. Phrase 5 nine measures); No, 26,
first 15 measures (two-measure Phrases, Group-formation of 2nd Period) ; No.

Par. 68b. THE EXTENSIONS OF THE DOUBLE PERIOD. Ill

32, first 14 measures (Introd. 4 measures ; 2-measure Phrases, Group-formation);
No. 36, first 27 measures; No. 39, first 15 measures (2-measure Phrases, a little
doubtful whether Single or Double-period form; Group-formation; Codetta
of I measure, repeated) ; No. 2, first 29 measures {seven Phrases, but unmistak-
ably Double-period form, parallel construction, enlarged by Group-formation
of 2nd Period).

Chopin, Preludes, op. 28, No. 3 (Introd. 2 measures; Phrase 2, five meas-
ures; Phrase 5, eight measures; Codetta); No. 4 (Phrase 2, eight measures;
Phrase 4,.five measures extended to nine); No. 5 (Introd. four measures; con-
struction parallel; Antecedent- group in each Period); No. 6 (Phrase 3, six
measures ; Phrase 4 reproduced), — Group-formation, as in Ex. 70 ; Codetta) ;
No. 13, first 20 measures. — Chopin, Mazurka 31 (op. 50, No. 2), first 28 meas-
ures (Introd. 8 measures).

Bach, Ouverture d, la mani^re fran9aise, ** Echo ", measures 1-22 (Phrase 2
eight measures; Phrase 4 also extended).

Schubert, Pfte. Son. No. 8 (C minor). Adagio^ measures 1-18 (4th Phrase
short); Pfte. Son. No. 10 (B|? major), ist movement, measures 1-18 (2nd
cadence expanded).

Brahms, op. 116, No. 3 (Phrase i of each Period repeated). — Op. 117, No. 2,
first 22 measures (Consequent-group at end). — Op. 119, No. i, measures 17-42
(parallel; cadence imperfect; each Consequent extended). — Op. 119, No. 3,
measures 1-24 (parallel; each Consequent extended).

(b) The device of transforming the (Large) Single period into
a Double period, by occasioning a semicadence in the center of each
half (par. 62, last clause), may also be applied to the Double period,
— which, if large enough, may thus assume the design of a

" C^ADRUPLE period".

See Mendelssohn, "Song without Words" No. 13, first 20 measures
(extended; it is not only possible, but perhaps necessary, to regard this as a
form of 4 — or 5 — four-measure Periods^ consisting of 2-measure Phrases).

Chopin, Mazurka 28 (op. 41, No. 3), first 38 measures; Chopin, Polonaise
No. 6 (op. 53), measures 17-48 (quasi repeated Double period) ; Chopin, Pre-
lude, op. 28, No. 16 (extended 13 measures at end).

Schubert, Pfte. Son. No. 5 (op. 143), first 46 measures.

EXERCISE 16.

Write a number of regular (i.e., unextended) Double periods, of Parallel
construction (par. 61-64).

Use both the major and the minor modes, best in alternate examples.
Employ different varieties of duple and triple time, and the different grades of
tempo from Adagio to Presto, And adopt different styles, imitating (even

112 THE EXTENSIONS OF THE DOUBLE PERIOD. Par. 68b.

closely or literally, if necessary) the style of certain given illustrations, or of
any other familiar or favorite composition.

See the n. b. on page 52.

Work towards the Cadence during each Phrase, and from Cadence i^
Cadence, Avoid vague forms of Cadence, as a rule, but employ freely the
varieties of concealed and evaded Cadence already learned and mastered.

A few illustrations of the Contrasting Double period may be attempted ;
best, as Large Single period with intermediate interruptions (par. 66).

EXERCISE 17.

The examples of Exercise 16, or new ones, extended in the manner indi -
cated in par. 67a, by c; and par. 68a.

par. Wc. THE SONG-FORMS, OR PART-FORMS. II3

DIVISION TWO.

THE SONG-FORMS,* OR PART-FORMS,

INTRODUCTORY.

69. The comparative definition of the various members and
distinctive elements of form which enter into the composition of
this second, larger, Division of structural evolution, is as follows :

(a) The Phrase: is an uninterrupted series of ** Chords",
and coherent melodic ' * Members " , of such a length as can be
reasonably sustained without palpable interruption, — generally from
2 to 8 measures ; closing with a Cadence (perfect Cadence or Semi-
cadence).

(b) The Part : is a coherent series of such ** Phrases *', to the
number of two, three, four or more ; (possibly only one ; and, on the
other hand, never more than can be held without effort in close re-
lation and connection with each other) ; terminating, as a rule, with
a strong Tonic Cadence.

(c) The Song-form : is a coherent series of such *' Parts", to
the number of 2, 3 or more; ending, as a rule, with the complete
Tonic perfect Cadence.

The cohesion of the members is observed to relax, in proportion to the
growth, or increase in length, of the form; simply because, the greater the
dimension, i. e., length, of the several sections of the design, the greater the
necessity of more emphatic points of repose, and of more variety and contrast
between the individual members. Hence :

Between the "Members" of the Phrase (i.e., the Chords and melodic
Motives) there is little or no separation (at most, a Quarter-cadence) ;
But between the ** Phrases " of a Part there stands a Semicadence;

* So called in analogy to the vocal forms from which they are derived ; but just
as common in instrumental as in vocal composition.

114 "^^^ TWO-PART SONG-FORM. Par. 7L

And between the ** Parts" of a Song-form there stands, as a rule, a com-
plete Tonic Cadence.

Furthermore :

The melodic and rhythmic elements of a Phrase must be verj closely re-
lated ;

The Phrases of a Part are organically related, but somewhat independent;

The Parts of a Song-form are still kindred, but sundered, and often quite
independent of each other, in certain outward respects.

70. According to par. 69^, the name '* Part "is given to any
series of Phrases, in comparatively unbroken succession, extending
up to a cadence of sufficient force to check both the harmonic and
melodic currents, and so to complete the musical purpose of th^t
section as to set it temporarily apart ; but not of sufficient culminat-
ing force to dispel the natural expectation of a following, kindre"d,
section, which may more fully consummate and confirm this musical
purpose.

(a) It is possible, as already hinted, that one single Phrase,
when of sufficiently opulent character, and especially when repeated^
might constitute an entire Part. But it is not probable ; and ex-
amples of so small a Part, excepting when occupying an interme-
diate position, between larger Parts, are rare. See Ex. 74 ; and
Ex. 78.

(b) As a rule, therefore, the '^ Part" will contain at least two
Phrases ; and may contain as many more as the limitation dictated
in par. 69^ will admit. In a word, each of the designs explained in
the jirst Division of this book^ — from the repeated Phrase up to the
enlarged Double period, — represents the elements, and is an example,
of the Single Part. And it is perfectly proper to speak of them
as the "One-Part" forms.

CHAPTER IX.

THE TWO-PART SONG-FORM.

71. In the Two-part Song- form, or, as it is also called, the
bipartite form, there are two such Parts ; usually, though not
necessarily, of corresponding general character and design, but effect-
ually disconnected from each other by a cadential interruption.

Par. 72c. THE TWO-PART SONG-FORM. II5

- 72a. Of these, the First Part, as already stated, may con-
sist of a Repeated Phrase (rare) ; a Period, regular or extended ; a
Double period, regular or extended ; or a Group of Phrases. It is,
perhaps, most commonly a Period, and regular in structure, «
keeping with the principle enunciated in par. 53 (which review).

The First Part of a Song-form is the ** statement'* of the leading musical
thought, — the motive, text, or subject; which, while it must be sufficiently
impressive, interesting, and pregnant to excite attention and give gratification
by itself, should depend chiefly upon the following Part (or Parts, if there be
more than two) for its development, elaboration and corroboration.

The First Part may be, and very commonly is, repeated.

(b) The Cadence of the First Part, as stated in par. 69^, will
be made, as a very general rule, upon some Tonic harmony, in its
strongest form, and in strongest rhythmic location. It may be (in
the order of preference) :

(i) The Tonic of the Dominant key, in a Song-form beginning
in major ;

(2) The Tonic of the Relative key, in a Song-form beginning
in minor ;

(3) The Tonic of the Original key itself.

(4) Or it may be the Tonic of some other Next-related key; i. e., the Dom-
inant from minor ^ or the Relative from major, — instead of the opposite, as
given above ; further, the Relative of the Dominant ; more rarely, the Relative
of the Subdominant key; and, most rarely, the Subdominant key itself; pos-
sibly some Remotely-related key, as the "Modulatory Stride",* or one of the
** Mediant '** keys.

(5) Other Cadence-conditions, to be avoided here, are explained in par. 93,
to which brief reference may be made.

(c)^The general modulatory current of the First Part will be
determinfed by the choice of Cadence. Review, in this connection ^
par. 37, recollecting that in a 2-Part form the First Part is an
"Antecedent ", simply magnified in dimension.

But, within this general modulatory design, there exists the
opportunity for transient modulations, more necessary here than in
smaller forms, because the increase in dimension magnifies the
simpler CAorfl?- relations into A^-associations. But see par. 72a,

*See the Author's ** Material used in Mus. Composition *', p. 149, and p. 155.

Il6 THE TWO-PART SONG-FORM. Par. 78c.

second clause ; and avoid overloading the Pirst Part with harmonic
and modulatory color. See also par. 94.

73a- The Second Part, in the 2-Part Song-form, is probably
most prone to assume the same form, or length, as its First Part,
though many digressions from this condition of symmetry are pos-
sible and common. In case of differentiation, the Second Part will
almost certainly be the longer of the two.

Like Part I, the Second Part may be repeated.

(b) In character, the Second Part must maintain fairly close
agreement with its First Part; not, by any means, servile thematic
relations (i. e., in respect of melodic design), but clo^e formative
connection (in respect of general harmonic character, rhythmic
character and technical style). As Part I is an ** Antecedent ",
Part II is its "Consequent"; therefore, a certain impression of
consistent opposition between the two Parts should be created, such
as distinguishes, on a smaller scale, the Period-relation. This is
very frequently effected by basing the first member of Part II upon
the Dominant harmony, in opposition to the Tonic basis upon which
Part I starts out.

In other words, while cadential separation is necessary, no radical
change of character should be perceptible in passing out of the First Part
into the Second. See par. 69, the very last clause. It is precisely such a dis-
tinct alteration of style ^ that ushers in a new ** Subject " or new ** Song-form "
(like the ** Trio " in the Minuet, — pars. 117, 119, — or the ** Subordinate Theme ''
in the Rondo and Sondita- Allegro forms). The end of the First ** Part " is suffici-
ently marked by the complete cadential break, demanded in par. 72^; but
the condition of coherency and consistency, uniformity of style and general
character, should prevail throughout all the component Parts of a Song-form.
On the other hand, as already stated, the thematic (melodic) conduct may
be as independent as is desired, and a certain amount of individuality is quite
essential.

(c) The Second Part ends with the complete Tonic perfect
Cadence, in the original key ; probably emphasized by a Codetta,
or, at least, by an extension at the end of the final Phrase. (See
par. 79^ and c. )

In its modulatory design, the Second Part will be found, in the
most perfect models of the 2-Part form, to incline toward the lower
(i. e., Subdominant) keys, — ^just as the First Part favors the higher
(i.e.. Dominant) direction.

Par. 74.

TWO-PART SONG-FORM, PRIMARY DESIGN.

117

Two-Part Song-form, Primary Dbsign.

74. The type of the primary 2 -Part Song- form (the design
out of which it is evolved) is the Double period, especially that of
contrasting construction, in which the continuity insisted upon in
par. 62 is disturbed by a cadential break in the center, so complete
that it resolves the * * One-Fa.rt ' ' Double-period form into a * ' 7w(?-
Part " Song-form. (See pars. 66, and 62.)

Hence, the Primary design of the 2 -Part Song-form is that in
which each Part is in the ordinary Period-form, with or without
repetition.

Examine the following illustrations of this Primary design scrupulously,
testing the minutest traits as they may be borne upon by the conditions
enumerated in paragraphs 71, 72 and 73 :

MotUrato,

71.

J '' i J./'j h^ ^ i J-;'j J

V

Ma JO It. Parti.

>artl. ♦!)

4 — i-

■xt

f

#

ff-

»■

U

i ^.JJ.li

f

EE

I I I I I II P»rtn. ^

j , J J J ,zt-d-T — i \-T-r I ! l J ■ jM I ! P \ \\

•4)
J

V *3)

2

-jBZ

'=A4^H-,-i-hH

SiciLiAM Hymn. (harm, by Stark )

f ^-^

' ; J J I I — I

^i

g^

»5)

■i9-

^

J

3z:

S^

1

ii8

TWO-PART SONG-FORM, PRIMARY DESIGN.

Par. 74.

*i) An unusually brief (light) Semicadence.

*2) This is a complete Tonic Cadence in the Dominant key, of such
weight and emphasis as totally to sever all purely external connection between
the two Periods, which, consequently, represent here two Parts, (Compare
par. 63^, last clause.)

*3) The Second Part starts out from the Dominant, while Part I began
upon a Tonic basis.

*4) The ** consistent opposition" of the Parts, mentioned in par. 73^, is
clearly illustrated here by the rhythmic form of this first measure, which cor-
responds to that of the second measure in Part I. Thus, while the rhythmic
figure of the very first measure of the Song-form becomes the rhythmic type of
Part I, that of the second measure characterizes the greater portion of Part
II; and the first two *^ measures^* constitute the rhythmic germ of the two
*' Parts ". This is consistent Musical Form.

*5) Here the Subdominant inclination of the Second Part (par. 73c) mani-
fests itself.

Andante sosUnuto,

72.

S

3 ^^
I ^^^

MiNOB, P

^3

^

m

JrJf l^^m

«1)

i

Part n. *2)

Par. 75.

TWO-PART SONG-FORM, PRIMARY DESIGN.

119

Mendelssohn. •S)

ft.

*i) A complete Tonic Cadence in the Relative key; see par. 72^ (2).

*2) In its thematic (melodic) aspect and style, this Second Part closely re-
sembles its First Part. But "consistent opposition" is nevertheless obtained
by enlarging the one-mesisure Sequence of Part I, to a /w^-m6asure Sequence
In Part II.

*3) Theme of the " Variations s6rieuses '\ op. 54. See also the individual
Variations^ which, while they nearly all preserve the 2-Part design, present
many instructive modifications in the treatment of the Cadences; viz., the
Cadence at the end of Part I is always bridged over, sometimes reduced to a
Semicadence, and is therefore more vague than in the Theme, — especially so
in Var. i, 2, 6 and 13; — Var. 3, 4, 7, 9, 11, 12 and 14 all approach the Double-
period form, because the middle Cadence is not strong enough to divide the
design into two Parts; — Var. 10, 15, 16 and 17 approach the form of a Group
of Phrases; — In Var. 5, 8 and 16, the middle Cadence is altered to a different
key; — In Var. 14, being Major, the Dominant (Semi-) Cadence is used.

See also: Chopin, Mazurka No. 50 (Peters ed.) first 32 measures (each
Part repeated).

Beethoven, 9 Variations in A major. Theme (by Paisiello), — Second
Part repeated. Beethoven, Symphony No. 2, Larghetto^ first 32 measures
(each Part repeated) .

Haydn, Pfte. Son. No. 4 (Cotta ed.). Largo (final Cadence on Dominant,
for an obvious reason) ; Pfte. Son. No. 5 (Cotta ed.), Presto^ first 16 measures
(two complete repetitions, the second one very much altered) ; Pfte. Son. No. 7
(Cotta ed.). Finale^ first 18 measures; Pfte. Son. No. 11 (Cotta ed.), first 16
measures; Symphony No. 9 (Peters ed.)^ Andante, first 10 measures; Sym-'
phony No. 6 (Peters ed.), Andante j first 32 measures (each Part repeated).

Schumann, ** Bunte Blatter*', op. 99, No. i (Cadence of Part I some-
what vague, — see par. 93) ; ** Waldscenen ", op. 82, No. 7 (" Vogel als Prophet ")
first 18 measures (Codetta of 2 measures).

Beethoven, Pfte. Son. op. 28, ** Trio ** of 3rd Movement (Part I repeated
literally; Part II repeated with modification; melodic relation between the
Parts unusually close); Pfte. Son. op. 57 (Appassionata), Andante, the Theme,
and its Variations also.

75. A very common, — almost characteristic, — trait of the
2-Part Song- form, exhibited in many examples in musical literature,

I20

TWO-PART SONG-FORM, PRIMARY DESIGN.

Par. 75»

consists in the sitnilariiy between the respective endings of the two
Parts.

Sometimes the resemblance is literal (as in Ex. 73) ; sometimes
the endings differ only in key (in case Part I does not close in the^
original key, — ^as Part II must) ; in some cases there is only a gen-
eral resemblance ; occasionally the coincidence embraces the entire
last Phrase of each Part ; and again, on the contrary, it is very
brief.

This point of resemblance between the endings of the Parts must not be
confounded with that essential condition of parallelism which prevails at the
beginning of each Period in the Double-period form (par. 64), for their respec-
tive influences upon the formal conception are precisely opposite. The agree-
ment at the beginning confirms and supports the coherency of the members ;
while the agreement of the endings, on the contrary, serves to emphasize their
separation^ inasmuch as the central cadence thus assumes a form and weight
similar to that of the final cadence, and closes its Part as completely as the
latter does the entire sentence.

N. B. Very particular care must be taken to limit any such re-
semblance as this strictly to the ending of the Parts, and to repress,
the tempting inclination to introduce, in the course of the Second
Part, any member which exactly corroborates the beginning of the
First Part I Such a feature would transform the imagined 2-Part
form into some variety of the Three-YzxX. design. See par. 8ia»

For example :

73.

AUegto,

tei^

V

^

^fe

li

-^--^-^

m

^ — «-

^%

'J y^ ^-

Part II.

TTlTy py -g^^

U t' 1 «i) ' *

Par. 76.

THE DIMINUTIVE TWO-PART SONG-FORM.

121

♦i) This is the comparatively rare termination of Part I with a perfect
Cadence in the original key ; see par. 72^ (3).

*2) T^e Thfeme of 8 Variations in Bj? by Beethoven (a pop ular Getman
song). The parallel endings are indicated by the brackets. See also the indi-
vidual Variations y and observe to what extent this structural feature is preserved
in some of them.

See also: Beethoven, 9 Variations in C minor (March by Dressler),
Theme, and all the Variations (resemblance covers the entire last Phrase, but
involves transposition).

Mozart, G minor Symphony, Finale, first 16 measures.

EXERCISE 18.

A. Write two examples of the 2-Part Song-form, Primary design ^ in
Major; with scrupulous regard to the details of structure and cadence given
in par. 72 and 73.

Use some variety of duple time for one, and triple time for the other; and
choose radically different style and tempo for each.

Review the directions given in Exercise 9.

B. Write two examples of the same form and design in Minor ^ observing
all the above directions ; in one instance introducing the similarity of ending,
explained in par. 75, and carefully guarding against the error pointed out in
the final clause.

The Diminutive Two-Part Song-form.

T6« As already declared in par. 70a, it is possible that a Part
may consist of no more than one single Phrase, and, though
naturally rare, such examples may be found.

This may be true of either Part, or even, though still more
rarely, of both. As usual, it is least likely to occur in the Second
Part (par. 73^5, last clause).

122

THE DIMINUTIVE TWO-PART SONG-FORM.

Par. 78.

How difficult it is to impart sufficient contents to a single Phrase^ and to
separate it so positively from its companion as to establish its dignity and in-
dependence as an adequate "Part**, the student will soon discover. Review
par. 57, second clause, and par. 70^; and reflect to what extent, on the other
hand, the continuity oi Phrase-succession . is maintained as One-Part form in
some of the given illustrations of the extended Double period, the Phrase-
Group, and the Quadruple period.

The conditions are most favorable when the Phrase, thus made
to represent an entire Part, is either of the large (8-measure) species,
or broad in tempo and character ; when it has a very firm Cadence ;
and when it is repeated. For illustration :

Andante,

74.

^

i

iN^

i

f

^

P

Part IT.

^

d:

^

i=i.

J

PP

Crete. ^

Par. 77. THE DIMINUTIVE TWO-PART SONG-FORM. 1 23

■^ Bbethovbn.

I I B » f f ^ i — i^-

^

*i)

*i) This is an unusual example of the 2-Part form, each Part of which is
only a 4-measure Phrase. But the breadth of character and contents, the com-
paratively moderate tempo, and the repetition of each Part, establish the form
beyond question. (There is even a remote possibility of assuming the Period-
form, instead of Phrase, in each Part ; compare par. 10.) It is to be found in
Beethoven's Pfte. Son. op. 27, No. i, first movement, and must he referred to.
The illustration here given covers the first 8 measures (or 9, with the second
ending); the following 12 (13) measures illustrate exactly the same variety of
the Diminutive 2-Part Song-form, the repetition of Part II being written out,
on account of unessential modifications ; the 16 measures which follow are a
recurrence of the above illustration, but with the repetitions written out and
modified; these are succeeded by an Allegro-Theme (also 2-Part form, but
" irregular," because the repetition of its Second Part is ** dissolved " — i. e.
conducted, as Re-transition, away from its own key into that of the next
Theitie) ; then follows another recurrence of the first Theme, with still other
varieties of modified repetition ; a Codetta of 8 measures concludes the move-
ment.

See also: Beethoven, Pfte. Son. op. 26, second movement, ** Trio "
(Part I an 8-measure Phrase; Part II a Period). — Pfte. Son. op. 79, Andante^
first 8 measures (Part I a Period, Part II a repeated Phrase) ; same Sonata,
Finale^ first 16 measures (Part II a repeated Phrase).

Beethoven, 10 Variations inB[7 (Salieri), Theme (Part I a repeated
Period, with peculiar transformation of the Dominant Semicadence into a
complete Cadence effect, by one measure of extension in the 2d ending;
Part II only a Phrase, but extended) ; see also each Variation ( — of Variation
10, only the first 45 measures). — Beethoven, Pfte. Son. op. 31, No. i. Rondo ^
first 32 measures (Part II a repeated Phrase ; the entire Song-form repeated
and variated).

Schumann, Jugend- Album, op. 68, No. 23 (Reitersttick) ; each Part
repeated literally, and then the entire Song-form repeated, with modifications ;
a long Coda follows.

Schubert Songs, ** Winterreise'% No. 5 (** Der Lindenbaum "), measures
9-24 (Part I is a repeated Phrase).

Carey, ** God save the King "—(Part I a 6-measure Phrase).

77. Reverting to par. 65, which review, attention must here again be
drawn to the perplexing, but altogether natural and essential, points of exter-
nal resemblance between different denominations of formal design. The

124 '^^^ DIMINUTIVE TWO-PART SONG-FORMS. Par. 77.

transitional stages from each one into the next higher, or larger, species are
so imperceptibly graded, that positive definition becomes, at a certain stage,
difficult and even impossible, — but also unnecessary.

Thus, it has already been seen (and in the student's subsequent personal
analysis of compositions, many further proofs will be encountered), that both
the Period-form and the 2-Part Song-form may consist of two Phrases ; that,
by inference, the extended Period, the Double period, the Phrase-Group (not
to mention the unusual designs of the Period-Group and Quadruple period)
might be greater^ in temporal dimension, than some specimens of the structur-
ally higher graded 2-Part Song-form ; that the external resemblance between
a Period in which each Phrase is repeated (Ex. 51), and a 2-Part Song-form in
which each Part is a repeated Phrase <Ex. 74), is complete, save in respect of
those few essential conditions upon which the distinction depends; that a
contrasting Double period can often scarcely be distinguished from a 2-Part
Song-form (see Beethoven, 24 Variations in A, Theme) ; and so forth.

The essential distinction between the ** One-Part" forms and the **Two-
(or Three-) Part " forms is clearly stated in par. 73^, second clause, and hinges
mainly upon the idea of Separation (which is as possible among kindred, as
among foreign members). Hence it is, that the repetition of either Part
serves to individualize it and separate it from its fellow ; that a sufficiently
pronounced difference in character, assumed perhaps abruptly, will have the
same effect ; and that, finally, simple emphasis of Cadence may suffice, even
in case of very close similarity of melody and style, to sever the Parts
unmistakably. See also par. 93.

EXERCISE 19.

A. Analyze the following (doubtful) examples, and endeavor to deter-
mine whether the form is Two-Part or One-Part, and, if the latter, exactly what
species ; pars. 93 and 104 may be referred to for additional clues :

Beethoven, Pfte. Son. op. 10, No. i, first 30 measures. — Pfte. Son. op.
10, No. 3, ** Trio'* of Menuetto. — Pfte. Son. op. 13, Finale^ first 17 measures. —
Pfte. Son. op. 31, No. i, first 30 measures. — Pfte. Son. op. 79, first 24
measures.

Mendelssohn, Caprice, op. 16, No. i, Andante section.

Schumann, Songs, op. 24 (** Liederkreis*'), No. 8.

Grieg, Lyric Pieces, op. 12, No. 2, measures 1-18.

B. Write two examples of the Diminutive 2-Part Song-form (par. 76),
one in major and one in minor; one in Adagio, the other in Allegretto tempo ;
one in duple, the other in triple time.

Par. 79b. THE FULLY DEVELOPED TWO-PART SONG-FORM. 1 25

CHAPTER X.

THE FULLY DEVELOPED TWO-PART SONG-FORM.

78. Here, all likelihood of a misconception of the form
(touched upon in par. 77) is precluded, because of the dimensions
generally reached, and the consequent opportunity, and even
necessity, of more distinctly individualizing and separating the two
Parts.

The conception of Part I will not differ in any other respect
from par. 72^, 3, c (which review) than in that it will surely never
be less than Period-design, and may be larger.

But the structure of Part II, while naturally adhering to the
conditions defined in par. 73a, ^, c (which review), will, in the
fully developed form, usually be more elaborate (longer, and with
more extensions) than in the smaller varieties explained in the
preceding chapter.

The designation ** Fully developed '* may be construed as applying to all
examples of the 2-Part Song-form in which each Part is at least a Period^ and
to some degree extended. Commonly, it is true, wider differentiation of the
Parts, in favor of greater length and independence of the Second Part, will be
observed.

79a. The EXTENSION of the Two-Part Song-form by repe-
tition of either Part, or of each Part, or of the entire Song, has
already been indicated in the text, and encountered in some of the
quotations.

The ** Variations *' upon a 2-Part Theme, cited above, may be regarded as
** modified repetitions '' of the entire Song-form ; in a broader sense, however,
than the pupil should undertake to exemplify for the present.

(b) Besides this means of enlargement and development (to
which may be added all the other ordinary varieties of extension in
the course of a Phrase, or of a section of any kind), there are also
those adjuncts of a more external nature, represented by

the Introductory Phrase,

the Prelude,
the Codetta or Coda, and

the Postlude.

126

THE FULLY DEVELOPED TWO-PART SONG-FORM. Par. 7«c.

(c) It is only necessary to remember that these auxiliary mem-
bers become more and more desirable, and are apt to assume
greater length, the larger the form to which they are attached.
Thus, the Introductory member of a 2- or 3-Part Song-form, will
probably, though by no means certainly, be longer, and perhaps
more independent, than that of a One-Part form.

Review pars. 44, 45; and see Mendelssohn, "Songs without Words",
No. 6, first 7 measures ; No. 23, first 6 measures. In both of these cases the
Introduction or Prelude takes a more active part in the plan of the whole, than
in some of the other ** Songs without Words " to which reference has been
made; and it will be observed that they recur at the end, as Codetta or
Postlude.

And, furthermore, the comparatively brief ** Codetta" or
Postlude added to a Period, or Double period (pars. 51, 52, 68a),
becomes sometimes a complete *' Coda" even in the 7w<?-Part Song-
form ; while in any still larger form this enlargement is all the
more probable.

The details of the distinction between Codetta and Coda are given at
length in par. 98, to which reference may be made, although the formation
and employment of the Codetta or Postlude must yet be restricted to the
directions hitherto given.

Illustration of the fully developed 2 -Part Song-form :

78.

AUfgro,

mp

^#1*: f at I ^^

jMr If

^A-trU

E

i

Par. 79c. THE FULLY DEVELOPED TWO-PART SONG-FORM.

127

I

marcato.

M

^

^^

■«^

p^ r^ i f iH-^^b^ ^

-^-

^^

#=p:

rrTT i iTrf i f

JoH. Seb. Bach. *1)

. I^^^_ J""' ^**' *5ACH. »'i;

r ar ' r tsl^ ff f

rctUerUando.

*i) Gavotte from the 6th French Suite. (The present author is respon-
sible for the phrasing and marks of expression.) Part I is. a regular parallel
Period, with perfect Cadence in the Dominant kej, according to the rule.
Part II is a Group of 3 Phrases, exhibiting the usual traits of established
form : the outset upon the Dominant harmony ; the inclination toward the
Subdominant key in the center, and again near the end ; and the general simi-
larity between the respective endings of the two Parts (more common, by the
way, in older than in modern music).

128 LARGE TWO-PART FORM, AS TYPE OF SONATINA-FORM. Par. 80.

See also: Beethoven, Bagatelle, op. 119, No. 8. — Beethoven, 8 Vari-
ations in F major (SCssmayr), Theme; 33 Variations in C major, op. 120
(DiABELLi) , Theme, and also the Variations (each Part a Double period ; very
similar endings).

Mozart, Pfte. Son. No. 3 (Cotta ed.), Andante^ first 20 measures
(design exactly like that of Ex. 75 ; the 16 measures which follow are a 2-Part
Song- form of Primary design).

Haydn, Pfte. Son. No. 8 (Cotta ed.), last Movement, ** Trio" (here Part
I is the larger design of the two, being a Group of 3 Phrases, while Part II is
Period-form ; but the number of measures is equal).

Schubert, Pfte. Son. No. 9, Andantino^ first 32 measures (an extraordi-
nary example ; Part I a Double period ; Part II almost identical tvith Part
I in its melody^ but harmonized chiefly in the relative key, and contracted, —
by which means it sufficiently establishes its identity as a separate Part; a com-
plete repetition^ with 4 measures of Plagal extension, follows).

Chopin, Mazurka No. 45 (op. 67, No. 4), first 32 measures (each Part
repeated. Part II literally; Second Part a Double period).

Chopin, Nocturne No. 10 (op. 32, No. 2), " Pii agitato^^ (V time;
Cadence of Part I very vague, but sufficient, under the circumstances; the
entire Song-form is sequentially reproduced^ a half-step higher, and led with-
out cadence into the following section) .

More elaborate and extended illustrations will be found in Mendelssohn,
Scherzo a capriccio in FJJ minor, first 36 measures (Part I a little longer than
II) ; the ensuing 10 measures are a Codetta, repeated and extended ; and the
next 32 measures are again in the 2-Part Song-form. — Mendelssohn^ ** Songs
without Words", No. 6, entire. — Mendelssohn, op. 72, No. 2, entire
(4-measure Prelude and Postlude). — Mendelssohn, fitude in a minor, op.
104, No. 3, entire (Part I an extended Double period — possibly Single period;
Part II Double period, or, more probably, a Group of 4 Phrases, No. 4
reproduced. Cadence somewhat vague; Coda 13)^ measures, to end).

Chopin, Mazurka No. 25 (op. 33, No. 4), entire section up to signature of
5 sharps (Part I, extended Double period, repeated; Part II, Period with
repeated Antecedent and extended Consequent; entire Song-form repeated
literally).

Brahms, op. 117, No. 3, first tempo (each Part a slightly expanded Period ;
entire Song-form repeated, with modifications).

Schubert, Impromptu, op. 90, No. 3, first 48 measures (Part II repeated);
also the following 53 measures (Part I has a 4-measure Codetta, repeated) .

Bach, Well-tempered Clavichord, Vol. I, Preludio No. 8 (£[7 minor), —
Part I, a group of 4 regular Phrases, resembling Double period ; Part II, a
Group of 3 Phrases, to which 2 others are added, after a- striking evasion of
the expected perfect Cadence (in measure 29) ; Codetta, 33^ measures.

The Large Two-Part form, as type of the Sonatina-form.

80. From the fully-developed 2-Part Song- form, of the
extent exhibited in the illustrations last cited, it is only one further

Par. 81a. THE THRKE-PART SONG-FORM. 1 29

step into that Large 2-Part Form out of which the Sonatina-
form emerges, and to which the latter probably owes its origin.
The demonstration of this Higher form must be reserved for a later
volume, as the student is not yet sufficiently equipped for its suc-
cessful manipulation. He is to make no other use of the Large
2-Part Song-form at present than to analyze the following ex-
amples. It will be observed that each Part is allied to the Phrase-
group form, usually with a Codetta ; and that the parallel structure
of the endings (cited in par. 75, — Ex. 73) generally extends over
the entire second half of each part,— often more.

See Haydn, Pfte. Son. No. 3 (Cotta ed.), Larghetto; Pfte. Son. No. 2
(Cotta ed.), Adagio; Pfte. Son. No. 11 (Cotta ed.), Adagio (a Coda of 15 or
16 measures is added). — Domenico Scarlatti (BOlow's revision, Peters ed.),
Suite I, "Sarabande" (Part I, extended Double period; Part II, group of
Phrases) ; same Suite, ** Burlesca '% " Menuetto '*, ** Gigue " (with brief Coda),
and " Toccata''. — Hummel, Pfte. Son. No. i (op. 2), Adagio (brief Coda). —
J. S. Bach, Well-tempered Clavichord, Vol. II, Preludio No. 15 (G major). —
Brahms, Intermezzo, op. 76, No. 3 ; this example is very diminutive, it is
true, but can scarcely be assigned to any other class of forms than that of the
"embryo Sonatina*'; (Part I, 10 measures. Codetta and Retransition 5
measures. Part II similar) ; it approaches the "Large Double period" with
Codetta to each Period.

EXERCISE 20.

A. Write two examples of the fully-developed 2-Part Song-form; one
of them in major and the other one in minor; one in Andante^ the other in
A ilegro tempo; one in duple, and the other in triple time; the second one
considerably larger than the first one. Add a Codetta to each. Observe the
directions given in par. 78.

For this work, one of the Periods invented in Exercise 9 might be utilized.

B. Write a third example, with free choice of all conditions ; but with
either an Introductory Phrase or a Prelude, and either a Codetta or a Post-
lude. — Again, some former Period-form may be used.

CHAPTER XI.

THE THREE-PART SONG-FORM.

8 1 a. The ruling principle of all 3-Part (or, as they are often
called, tripartite) Forms is, the Return to the Beginnings or to the
first (or principal) thematic section.

130 THE THREE-PART SONG-FORM. Par. 81c.

In order to be an unmistakable and perfectly genuine Return,
it can only succeed an equally genuine and positive Departure ;
and this Departure is embodied in the correspondingly distinct
Second Part, which intervenes between Part I (the definite
Statement) and Part III (the Recurrence of the Statement).

(b) This specific return to the beginning, or, more properly, this recur-
rence of the first thematic member as a distinctive trait of the 3-Part forms,
must not be confounded with those examples of apparently similar recurrence
that have been seen in repetitions, and in the Period and Double period of
parallel construction. For instance, in Ex. 25 the initial melodic member is
observed to recur in the 5th measure ; but this initial member is connected
with all that follows during 4 measures, exactly as in the preceding 4 measures,
so that the contents of measures 5, etc., are an immediate repetition without
any intervening digression or definite Departure. And even in Ex. 31, where
there is an intermediate passage between the first Phrase and its recurrence as
repetition, it is merely an "Interlude,'' with none of the elements of a
** Departure." See the last clause of par. 20.

In Ex. 46 the recurrence, in measures 5, etc., of the first members (con-
tained in measures i to 4) is again an immediate reproduction, not separated
by an intermediate Departure. The same is the case in Ex. 67, though there
is much more reason for misconception here than in the smaller examples,
because something really appears to intervene between the first melodic
member and its recurrence in measures 8, etc. ; but this intervening section,
while it is an entire Phrase (Phrase 2 of the parallel Double period), is un-
interruptedly connected with the first Phrase, and does not constitute that
degree of ** genuine and positive Departure" upon which the "Return"
depends, as distinctive condition of the 3-Part forms. Here, again, it is
simply a question whether the form is only " One-Part," or more; and the
decision, as shown in par. 77, depends upon the degree of separation, which
either is, or is not, sufficient to define a genuine Departure. Comparison of
the above example (67) with Ex. 76, where the interruption is complete, will
make this plain.

(c) The influence which this leading principle exerts upon the
^conception of the Second Part, is obvious. In the 2-Part forms,
Part II was found to be a coordinate companion to Part I, carrying
on the development of the melodic purpose with preponderant
parallelism of design, as ** continuation " of the musical sentence;
and keeping the perfect cadence of the original key in view, as
final aim.

But in the 3-Part forms, on the contrary, Part II is not so much
a continuation as it is a ** digression" ; it is not coordinate with
Part I, and the idea of opposition will be more likely to predom-
inate than that of parallelism ; it is not a final, but an inter-

J

Par. 83a. THE THREE-PART PERIOD. I3I

mediate section of the form, and does not tend to a point of rest,
but strives (from the moment when its identity as *' Departure"
has been assured) to regain, or lead back into, the original melodic
current, i. e., to prepare for the recurrence of Part I. See pars. 87
and 88.

The Three-Part Period.

82. The embryo of the 3-Part Song-form is a diminutive
design, which embodies all the essential conditions of the fully
developed form, and for which the term Three-Part Period
seems most appropriate.*

It bears the same proportion to the full 3-Part Song- form, that
Ex. 74 does to the full 2-Part Song-form, inasmuch as each of its
three sections is only a Phrase. The feasibility of this diminutive
design is indirectly demonstrated in par. 76, w^hich review.

The designation ** Period" instead of ** Song-form", though somewhat
inconsistent, is justified by the resemblance of this form to the different varie-
ties of Periods extended to the sum of three Phrases (i. e., Period with Conse-
quent repeated, or Consequent-group; Period with Antecedent repeated, or
Antecedent-group), in each of which a reduction of the three Phrases to two is
possible. The impossibility of such a reduction in the "Phrase-group" is
the very reason why the liatter could not be spoken of as a ** Period " (Ex. 63,
note *4)) .

83a.. In the genuine 3-Part Period (as type of the corre-
sponding Song-form) the First Phrase ends with a perfect cadence^
either in the original key, or in some next-related key ; and it is
generally repeated.

The Second Phrase ends, as a rule, upon the Dominant har-
mony, because that is the most favorable point from which to regain
the Tonic harmony with which the first Phrase (presumably) began.
Phrase II is not repeated alone, as a rule.

The Third Phrase is a more or less exact reproduction or re-
currence of Phrase I. At least \\\q first member should thematically
corroborate that of the first Phrase. Phrase III is not repeated
alone, hut Phrases II and III are frequently repeated together.

For example :.

*The designation *' Diminutive 3-Part Song-form " might seem more appropriate thftn
" 3-Part Period ", from analogy with par. 76. But the latter is, for valid reasons, neyer*
theless the preferable term. See Ex. 76, Note *x).

132

THE THRKK-PART PERIOD.

Par. 88a.

76.

Phrase II.

I

i?

]}\> : V

e^

k

^±±:

i»

^^

/P

•8)

i

?»^

!»

m

T

:j-^

|5t 5

*i) The weight of this perfect cadence is sufficient to detach the Phrase
positively from those which follow, and to make it as completely a Part as is
the case with the first Phrase in Ex. 74. This betrays the sole comparative
inconsistency of the term 3-Part Period^ to which reference has been made.

*2) The cadence of Phrase II is made at this point, upon the Dominant
harmony, and is so bridged over as to lead without the slightest check (almost
like an Introduction) into the initial motive of Phrase I, — now Phrase III.

Par. 83b.

THE THREE-PART PERIOD.

133

*3) The first ending of Phrase III (which is a literal recurrence of Phrase I)
is as conclusive as at first ;

*4) but in the second ending the perfect cadence is evaded, and spun out,
in Chain-phrase formation, into what, for convenience, may be called a Codetta.

See also: Schumann, "Album-Blatter", op. 124, No. 2 and No. 9 (con-
taining all the repetitions).

Schumann, "Jugend- Album" op. 68, No. i (all the repetitions) ; No. 10,
F major (Phrase II only 2 measures; all repetitions); No. 17, A major
(Phrase III modified and extended ; Codetta, 9 measures).

Beethoven, Bagatelle, op. 33, No. 4, first 16 measures (all legitimate
repetitions) ; Bagatelle, op. 33, No. 6, first 20 measures (Phrase III is repeated
alone, like Phrase I, before the repetition of Phrases II and III together begins).

Beethoven, Pfte. Son. op. 2, No. 2, the "Trio" of the 3rd movement
(8-measure Phrases) .

Haydn, Pfte. Son. No. 12 (Cotta ed.), Presto^ measures 25 — ^44 (Phrase I
repeated ; Phrase III modified, and extended to 8 measures).

Chopin, Nocturne No. 16 (op. 55, No. 2) first 12 measures (a superb illustra-
tion; Phrase II closes with decided cadence upon the Dominant of the relative
key, and the remainder of the measure, — the 8th, — is retransitional ** bridging "
of the cadence-measure ; the first measure of Phrase III is a somewhat disguised
recurrence of the beginning, but unmistakably the same member ; and the rest
of the Phrase is a correspondingly ingenious ornamentation of Phrase I).

Chopin, Mazurka No. 4 (op. 6, No. 4), each Phrase repeated — even the
Second one ; Chopin, Mazurka No. 9 (op. 7, No. 5), each Phrase repeated;
Phrase II a simple transposition of Phrase I into the Dominant key ; Phrase
III a literal "da capo," and, as expressly stated, running on "without end."

(b) Other, more irregular, varieties of the 3-Part Period afTord
perhaps greater justification of the denomination "Period,'' but
resemble the genuine 3-Part forms all the less. In all of these cases,
it is true, the Third Phrase is substantially the same as the First
one, but the condition of the cadences does not conform to the
rules given in par. 83a. For illustration :

Moderaio.

Phrase III.

w Phrase J

*2) \j V

Folk-song.

/

134 THE THRKE-PART PERIOD. Par. 8Sb.

*i) The First Phrase closes, irregularly, with an imperfect cadence (the
chord-third in Soprano) ; but it is of ample weight to define the termination of
its Phrase, and it will be observed that the same imperfect cadence recurs
at the very end, — a peculiarity of certain Folk-melodies.

*2) This cadence is exactly the same as that of Phrase I ; therefore it
furnishes no sufficient proof of the qualities which identify the ** Second
Phrase *' of a tripartite form, and even seriously disturbs the coherent connec-
tion with, and impulse inio^ the Third Phrase (par. 83a). But, notwithstanding
this default in its preparation, Phrase III is clearly recognizable as a Recurrence
of Phrase I, chiefly because of its distinct rhythmic character.

See also : Schumann, ** Jugend-Album '* op. 68, No. 19, a minor (Phrase I
closes with a Dominant semicadence; Phrase II is only 2 measures long; the
beginning of Phr. Ill is disguised; all the legitimate repetitions occur); No.
38 — '* Winterzeit, I." — (Phrase I ends yi\\i}[i 2i semicadence ; Phr. Ill considerably
altered in its second half; all repetitions). — Schumann, "Album-Blatter" op.
124, No. 13 (instead of. repetition. Phrase I is reproduced as sequence; Phr. II
closes with the same imperfect Tt>nic cadence seen in Example 77 ; Phr. Ill is
an extended Recurrence of Phrase I). In this case the designation Phrase-
Group would be admissible, and doubtless even more accurate.

Chopin, Nocturne No. 7 (op. 27, No. i), first 26 (28) measures. Each
Phrase is repeated, even the Second one, alone. The repetition of Phr. I
closes with an expansion of the penultimate cad. -chord, and is interlocked
with the beginning of Phr. II by an Elision ; Phr. II, and its repetition, end
with a full cad. upon the Tonic of the original key ! This almost totally severs
its connection with Phrase III, which consequently, though an almost exact
reproduction of Phr. I, assumes rather the character of a Coda. To this
example, also, the conditions of the Group of Phrases^ given in par. 58, appear
to apply more exactly than do those of the 3-Part Period.

Brahms, op. 118, No. 4, first 16 measures (closes with semicadence; Phr. 2
extended).

These more irregular examples all point to the fact that the similarity
between Phrase III and Phrase I, while corroborating the ruling principle of
the 3-Part designs, is not enough, in itself^ to establish the legitimate "3-Part
form." For instance, this very likeness appears plainly in the parallel
** Double period," as emphasized in par. 81^. It is only when associated with
other characteristic conditions^ that this structural trait becomes a perfectly
valid factor of the tripartite design. And it must be remembered, that it is
only because the absence of these other conditions is more easily accounted for
and excused in smaller designs, where quick comparison is possible, that such
irregular and almost indefinable forms as those last cited can exist. In the
fully developed 3-Part forms there is far less likelihood of such confusing
irregularities, and therefore the ** Return to the beginning" may be accepted
as an unfailing proof of the identity of the forms under present consideration.

EXERCISE 2L

A. Write two examples of the regular 3-Part Period, one in major and
one in minor; choosing a different time, tempo and character for each ; and
adhering strictly to the directions given in par. 83<i.

Par. 84b. THE INCIPIENT GRADE OF THE 3-PART SONG-FORM. 1 35

For this exercise, choice may be made among the Phrases invented in
Exercises i and 2.

B. Write a third example, with modified repetitions as prescribed in the
legitimate designs ; and with a partial change in the formation of Phrase III
(as compared with Phrase I); also add a Codetta.

The Incipient Grade of the 3-Part Song-Form.

84a- In this design the First Part is at least a full Period^
while Parts II and III each still adhere to the diminutive form
of the Phrase. The term *' Song-form" must be adopted, because,
as soon as any one of the three sections assumes the Period-form, it
becomes a full ** Part," and it is no longer consistent to speak of the
whole as a 3-Part ** Period."

(b) In its details, the Incipient 3-Part Song- form corresponds
to the schedule given in par. 83^5, only excepting the enlargement of
Part I, cited above ; namely :

Part I is a Period, of any variety, (possibly, though rarely,
anything larger,) and with a complete Tonic cadence in the original
key, or in the Dominant key (if major), or Relative key (if minor),
or perhaps in some other related key.

Part II is only a Phrase (possibly extended), with a very definite
Dominant cadence that strongly suggests, and leads into, the first
thematic member of the First Part.

Part III also is only a Phrase ; therefore it is not a complete
Recurrence of Part /, and discrimination must be exercised, that,
in thus abbreviating the contents of the First Part, the ruling con-
dition of tripartite form be not sacrificed : Under all circumstances
Part III must^ at its beginnings distinctly confirm the beginning
OF Part I. If Part I be a Period of parallel construction. Part III
may appear to be a recurrence of the second Phrase of Part I,
inasmuch as the beginning of that Phrase will then correspond
to the beginning of the whole. But if Part I be a period of con-
trasting construction, then Part III must corroborate the frst Phrase
(partly or entirely), and may not resemble the second Phrase at all.
It is not necessary to specify just how much of Part III may or must
be derived from the First Part ; but, as a rule, at least one full
measure, or one complete melodic member, should agree with the
initial measure or member of Part I. The remainder of Part III
will be" dictated by the course of the "highway to the perfect
cadence " (defined in paragraphs 6 and 7). For example :

136 THE INCIPIENT GRADE OF THE 3-PART SONG-FORM. Par. 84b.

GraMtbsc. • •

Part I. (Period).

78.

• •

A il I I ■ , ^3j j,-..^^

♦£ t

^j=f&?f^

If

«=t

w=r.

TTTT i Qn

3

^-TE-^-

^

Part II. (Phrase).

II

^

ffi

?

«=t:

t

'Fi=B^

£XJ

1

JlSS^

r

^

Pan III.
(Phrase).

*?:£ t

Par. 84b. THE INCIPIENT GRADE OF THE 3-PART SONG-FORM.

137

Bbethovbn. *1)

*i) Pianof. Sonata op. 2, No. 2, Finale, Part I is a parallel Period, with
a very striking initial member, the recurrence of which is therefore quickly-
recognizable. It closes with a perf. cad. in the Dominant key. The Second
Part is in reality only a 2-measure Phrase, made four by repetition. Part III
appears, at first glance, to be a recurrence of the second Phrase of Part I,
but on closer examination it proves to be much more nearly identical with the
first Phrase, — excepting, of course, that it terminates with the regular perfect
cadence. But even if it did resemble the 2nd Phrase more than the ist, the
** Return to the initial member'' would be unmistakably defined; and the two
Phrases of Part I, constituting a parallel Period, represent in total substance
but little more than one Phrase, anyway.

See Beethoven, the same Sonata (op. 2, No. 2), Largo, the first 19
measures. Here the derivation of Part III from the first Phrase of Part I
is indisputable, for the construction of the latter is chiefly contrasting. Part II
ends with the usual Dominant cadence, and the cadence-measure is so bridged
over as to lead very smoothly into the initial member. Part III is extended to
7 measures, but remains a ** Phrase " in form.

In Beethoven, Pfte. Son. op. 7, Finale, first 16 measures, the resemblance
of Part III to the second Phrase of Part I is stronger. But there is no doubt
that the impression conveyed is that of a ** Return to the initial member."
With this idea established, there is no danger of confounding the Incipient
S-Part form 'with that variety of the 2- Part form in which the Parts have
only a similar ending. See Ex. 73, and the N", B. immediately preceding it.

See further: Beethoven, Pfte. Son. op. 14, No. 2, Andante, first 16 meas-
ures (Part III corroborates the first Phrase of Part I, excepting the Cadence;
a Codetta of 4 measures follows). — Beethoven, Pfte. Son. op. 49, No. i,
Finale, first 16 measures (Part III is an almost exact copy of the second Phrase
of Part I, but the construction of the latter is parallel). — Beethoven, Pfte.
Son. op. 2, No. I, Adagio, first 16 measures (a misleading example, open to
difference of opinion; Part III is the corroboration of the initial member,
in a disguised — though by no means unrecognizable — form). — A still more
questionable example is Beethoven, Pfte. Son. op. 22, minor e of the Menuetto ;
this is very probably 2-Part Song-form. — Further, Beethoven, Bagatelle,
op. 33, No. 3, first 16 measures (a very clear illustration); Bagatelle, op. 119,
No. 4. .

Schubert, Impromptu, op. 142, No. 3, Theme.

138 THE INCIPIENT GRADE OF THE 3-PART SONG-FORM. Par. 84b.

Mozart, Pfte. Son. 15 (Cotta ed.), Theme of last Movement; see also the
several variations. — Mozart, Pfte. Son. 9 (Cotta), Theme of first Movement ;
here Part III almost exactly resembles the second Phrase of Part I — with an
extension of 2 measures, — ^and represents, therefore, what might be regarded
as the more questionable class of Incipient 3-Part Song-forms. What Mozart's
own conception of the "Idea" was, is however manifested in the succeeding
variations, in all but one of which the resemblance of Part III to the first
Phrase is very close, to the exclusion of all but an allusion to the style of the
second Phrase, in the added Extension. — Mozart, Fantasie and Sonata in
C minor (Cotta ed. No. 18), Andantino, entire; all the legitimate repetitions
are made, but written out and variated.

Haydn, Pfte. Son. No. 4 (Cotta ed.). Finale, measures 21-40 (Parts II and
III each extended to 6 measures) ; Pfte. Son. No. 8 (Cotta), Scherzando, first
16 measures.

Grieg, Lyric Pieces op. 12, No. 3, first 16 measures (followed by repetition
of Parts II and III). — Grieo, Ballade op. 24, Theme.

Chopin, Nocturne No. i (op. 9, No. i) measures 19-50; this example,
(which contains all the repetitions), is again, on account of far more striking
likeness of Part III to the second Phrase of Part I, one of the doubtful class;
the evidence of a ** Return to the beginning*' is, however, sufficiently striking
to define the tripartite design. The same is true of Nocturne 10 (op. 32, No. 2),
first 18 measures (2-measure Prelude). In par. 100 may be found another
possible classification of these two, and all similar, examples.*

*The classification of Ex. 78 (and the illustrations in the additional refer-
ences) among the 3-Part forms, differs from the analysis hitherto adopted by
other writers, who rank them among the Two-V?iTt forms. Though this is not
the only question upon which the present author holds a different opinion
from that of other theorists, it is the only one to which a few words of defence
are devoted. It appears obvious, from the great diversity of dimensions that
the pupil has already observed to exist within one and the same structural
design, and from the confusing similarity of size often attending entirely dif-
ferent forms, that dimension and proportion alone cannot be reliable criteria,
and that, therefore, the classification of the various designs must depend rather
upon the idea embodied in them. For this reason a distinction appears to be
necessary between the Sections of a form (which are the more mechanical
divisions), and the Parts (which are the ideal divisions). In every 3-Part
Song-forni a division into two sections will be observed, the first of which agrees
with the first Part, while the 2nd section embraces both Parts II and III, which
are so inter-dependent in the ** idea" of the tripartite form, that their separa-
tion would violate its chief condition. The mere accident that, in the above
Example (78), this 2nd section, — Parts II and III, — is precisely the same size
as the first section, is not sufficient proof of the 2-Part form; for ** Section "
and " Part" are not the same unless they are identical in the idea they embody.
If it be contended that the single Phrase is not enough to represent a Second
Part, proof to the contrary may be found in Beethoven, Pfte. Son. op. 49,
No. 2, Finale^ first 20 measures; Haydn, Pfte. Son. No 12 (Cotta ed.). Finale,
first 24 measures ; each of these is a full 3-Part Song-torm, in which Part II is

Par. 85. THE ORDINARY COMPLETE 3-PART SONG-FORM. 1 39

(c) How easily the Incipient 3-Part Song-form may be evolved
out of the 3-Part Period, is seen in such examples as Beethoven,
Bagatelle op. 33, Nos. 4 and 6 (both cited after Ex. 76 as illustra-
tions of the 3-Part Period), in which the modified repetition of
the Phrase that constitutes Part I, is so seductively suggestive of the
parallel Period; especially in No. 6, where Part III reproduces
this repetition; and still more so in Bagatelle op. 33, No. i, in
which, besides this same misleading trait of Parts I and III, the
Second Part, while only a Phrase, is extended by repetition and
other means to the length of no less than 16 measures. Thus these
examples, while actually only 3-Part Periods, assume the external
appearance and dimensions of the broader Song-form.

BXBRCISB 22.

Write two examples of the Incipient 3-Part Song-form ; one in major and
the other in minor; observing the directions given in par. 84^.

In the first one, construct Part I as parallel Period; in the other, as
contrasting Period.

CHAPTER XII.

THE ORDINARY COMPLETE 3.PART SONG-FORM.

85. In this species, Part I is at least Period-form; Part II
generally also at least a Period ; and Part III is a nearly or quite
literal Recurrence of Part I. The details of the design are as
follows :

only a Phrase, but nevertheless a perfectly distinct ** Departure.'* The idea
of all bipartite forms is solely that of thesis and antithesis ; the simple opposi-
tion of Question and Reply, or Statement and Counter-statement, be it upon
parallel or upon contrasting lines. This has been found thus far in the Period,
Double period, and in the 2-Part Song-forms exhibited in Exs. 71, 72, 73, 74
and 75. In Ex. 78, though it corresponds in external appearance to Ex. 71, 72
and 73, there is something more than this; there is the three-loXd idea of
Statement, Departure and Recurrence; on a smaller scale, it is true, than
in the examples which are to follow, but just as genuine and unmistakable as
in any full-fledged 3-Part Song-form, or Rondo, or even Sonata-a//«'^ro, for
that matter. The presence of this idea has led the present author to classify
these examples among the 3-Part forms, with no more than the just reserva-
tion indicated by the term "Incipient Grade''; and his experience, during
many years' teaching, of the readiness of the pupil to grasp and apply this
distinction in the work of analysis, testifies to its convenience and to more
absolute reliability than can be secured in many other, far more perplexing,
phases of Form-evolution.

140

THEMATIC CONDITIONS OF PART II.

Par, 87a.

The First Part.

86a.. The First Part of the tripartite form will not differ
in any essential respect from the First Part of a bipartite design.
Therefore the details given in par. 72a, ^, and c apply here, and
must be carefully reviewed.

(b) The length of Part I should be taken into consideration, and deter-
mined if possible beforehand, in proportion to the probable, or intended,
extent of the entire Song- form. As alreadj stated, Part I should be, as a
rule, a briefs simple^ clear Statement, A long First Part involves either the
necessity of extreme length of the whole Song-form, or the defect of top-
heaviness. A glance at Mendelssohn, "Song without Words'' No. 9, and
then No. 10, makes this sufficiently plain.

The Second Part ; thematic conditions.

87. On the contrary, the Second Part of a tripartite design
differs essentially from Part II of a bipartite form. See par. 8ic.
The distinction tends to lower the rank, melodic importance and
structural independence of the Second Part, but by no means
to the exclusion of striking individual qualities.

Close FORMATIVE RELATION to its First Part (in regard to
style and general character) should be strictly maintained. Review
par. 73^. But considerable freedom may be exercised in its melodic
delineation, which may be suggested entirely, or in part, by the
melodic design of Part I, or may be entirely independent of the latter,
especially when similarity of style and character is preserved.

Illustrations of various phases of thematic relation follow : —

(a) First, almost total agreement between Parts II and I :

Con moto.

7a rjAyJ.|J;'j ;jjjj.r^^^

Parti. Phrase 1.

i

fc

s

-^-

^

5

Phrases.

N.B. ,

l- y> jjti4 ^ ;j /jjj i .fr |^i|- ii J'*-iJi»^

Phrase 3.

EjAb.y tj. i ^f^;.i4^i^

Phrase!

^

\"1

Par. 87b.

THEMATIC CONDITIONS OF PART II.

141

m

?=^=^

g

&^^

Phrases.

Part II. Phrase 1.

±=t

±^tj=t

^-r

*

i^^SE

Phrase 2.

i

W-

m

-^-

^

**

£

-^-H '" "^f

Phrase 3.
Mendelssohn. *l)

etc.

*i) ** Song without Words" No. 13; see Original. The ist Phrase of
Part II is an exact transposition of Phr. i of the First Part; the 2nd Phrase
(Part II) is borrowed similarly from Phr. 5 of Part I ; Phrase 3 (Part II) is
also derived from the latter, but contains, besides, a new and rhythmically
characteristic member. — The Second Part is simply a Sequence, or transposed
reproduction, of Part I, in Schumann, Jugend-Album, op. 68, Nos. 3 and 8;
and mainly the same in Beethoven, Pfte. Son. op. 26, Marcia funebre. — In
Schubert, Impromptu op. 142, No. 2, **Trio," the Second Part corresponds
almost exactly to Part I, for a few measures, but transposed to the minor mode.
— Chopin, Mazurka No. 51, meas. 1-32. — In Mendelssohn, ** Song without
Words'' No. 42, the melody of Part I is reproduced in Part II, Phrase for
Phrase, in variously transposed and somewhat modified form, and is extended
by the reproduction of the last Phrase.* — See also No. 12, in which, again,
every figure of Part II can be traced back to the First Part. — Also Nos. 39, 36,
20, and 10, where the thematic derivation of Part II from Part I can easily be
traced. See also Ex. 72, note *2).

(b) Secondly, Part II derived from secondary members of the
First Part :

1. Parti.

*In using these references, and those which follow in this paragraph, the
pupil must examine no more at present than Parts I and II, and only those
points of the latter which reveal its thematic relation to the First Part.

142

THEMATIC CONDITIONS OF PART II.

Par. 87b.

Pait II.

NB.

a. Parti.

^y^s- u^

Part II.

*i) Mendelssohn, "Song without Words '' No. 14. See the Original.
The thematic derivation of the Second Part from the First is indicated by the
brackets. It will be observed that the second member of Part I becomes
the Jirst (and therefore typical) member of Part II; as in Ex. 71, note *4),
which review.

*2) *• Song without Words '* No. 29, — same as note *i). Here the deriva-
tion is cleverly disguised, but not sufficiently to arouse doubt of the composer^s
intention (or perhaps partly unconscious conception), in the mind of the closely
observant student. This is one of the ** secrets " of the true spirit of classical
Form.

See also : ** Song without Words '' No. 21 (Part I is repeated ; Part II starts
out, in meas. 61, with measure 11 of the repetition of Part I) ; No. 28 (Part II
based largely upon the last figure in the first Phrase of Part I) ; No. 25 (Part
II utilizes at first melodic fragments of the first and 3rd measures of Part I, —
to the exclusion of the first figure, in the preliminary half-measure ; after
a while this figure gradually reasserts itself, and finally the entire first Phrase
of Part I is taken up twice in Part II, of course in a different key). See also
Mendelssohn, op. 72, No. i (Part II starts out with measure 5 of the First
Part) ; op. 72, No. 3 (the second half of measure i becomes the first half, in
Part. II); op. 72, No. 5, — like note *i) above.

Par. 87b.

THEMATIC CONDITIONS OF PART II.

H3

Sometimes Part II begins where (so to speak) Part I leaves off :

Allegro.

81.

Parti.

^m

^^

i

fe

b> X

fe

TTTTTT . ^ g^

SE i ji ^ ^

4 — ♦— — h-

s:

Partn.

p *1)

*i) This agrees with the last member of Part I (as found in the lower
parts), but with nothing else that the Part contained.

*2) This figure corresponds to measure 6 of the First Part, in opposite
direction (** contrary motion " ; see Ex. 82). — See also Ex. 89; also

Haydn, Symphony No. 6 (Peters ed.), Menuetto (Part II is built entirely
upon the last mel. member of the First Part); Symphony No. 9 (Peters ed.),
Menuetto (Part II derived from last figure, in the Codetta to Part I).

144

THEMATIC CONDITIONS OF PART II.

Par. 87d.

Schubert, Pfte. Son. 4 (op. 122), Menuetto; Son. 6 (op. 147), Scherzo.
Schumann, Papillons, op. 2, No. i.

(c) Thirdly, Part II constructed more or less in the direction
opposite to that of the First Part, — similar to par. 39^ ; especially
at the start :

I nt J etc.

etc.
Haydn.

t=t:

b ^= h^^^^

a. Part I.

-M

h

t=i

1

Part II.

etc.

Haydn.

Mbndblssohk. No. 46.

Part I.

Part II.

See also Ex. 87. — Mendelssohn, ** Song without Words" No. 9 (the first
figure is in '* contrary motion**; the remainder chiefly parallel); No. 8 (the
same); No. 30.

Brahms, op. 76, No. 4. — Schumann, Jugend-Album op. 68, No. 12, first 24
measures.

This thematic relation (or opposition) of Part II to Part I is one of the
characteristic features of the old ** Gigue-form,*' whether constructed in 2-Part
or in 3-Part Song-form. See Bach, English Suites : No. 3, Gigue ; No. 5,
Gigue ; No. 6, Gigue ; ** Well-tempered Clavichord " Vol. II, Prelude No. 5.

(d) Fourthly, Part II melodically different from the First Part,
though similar or related in general style ; the same vein of melodic
conception simply runs on over into the Second Part, which, in this
variety of the form, becomes thus more nearly coordinate with (less
dependent upon) Part I :

Par. 87d.

THEMATIC CONDITIONS OF PART II.

HS

AndanU»

ms^

* — #-

^s^

m

I

fe

1*

V

whtm

^

ir

r r^^

ft;

j7

^

f

!»

1i

^

I

i*

t

F

♦1) Part XL

Chopin.

^>> •' J J '^;.3 i j . ^^m

^m

^

^ ^ -^

riien.

f

■«— ;^4 «•

f =-^' f^ ^* If '

^M

etc.
Partm.

i

146

THEMATIC CONDITIONS OF PART II.

Par. 87e.

♦i) In the Original (Nocturne 15, op. 55, No. i) the repetition of Part I is
written out and very slightly modified.

*2) The organic continuity of the Parts is upheld by similarity of general
style, and uniformity of accompaniment.

*3) This is the only place in Part II where real thematic agreement with
the First Part can be detected, though the resemblance between this and the
very first member is probably only accidental.

See also : Mendelssohn, ** Songs without Words," Nos. 37 ; i ; 40.; 16; 4,

Schubert, Pfte. Son. No. i (op. 42), "Trio" of 3rd Movement.

(e) Finally, Part II may diverge still more widely from Part
I, and be not only thematically *'new," but even somewhat inde-
pendent in character and style. But see par. 73^. For illustration :

CJuer/ulfy,

^ J r ' i l « '-^^

_> '^.

Parti.

"*/

fi:

g^g=s

±^

tf^Jii

*=t^

jO^y f l\

^^^^

-N— ,

?

i

t

m

r F f ,_!bba^^=^-' :

^==;?^

PartlL ♦!)

\ ^

\P ^ P^

^

y

i^^^^^r^i

w^

ft

£=^=5

i:

u V V ]/ ^ ^

P : i7 f -L L- 1 jJJ-^-tT^—'frh^

f p

^=f=F^=F

'^m

:t

■^F— »-

M

Par. 88.

TONALITY OF PART II.

H7

^'^ 'i V rt

i

y

it

vf-

T[

53^

y~r

J:.

Schumann.
Partin.

^

T

i

i

y

^— I

>u

^F — »

^

P=p:

^£

i

*

9^:

t

*i) Certain slight thematic coincidences exist between this Second Part
and its First Part, but the impression conveyed is that of a more pronounced
change of character than is exhibited in the examples preceding; though
the difference between this stage and that of Ex. 83 may appear scarcely
definable.

See Mendelssohn, ** Songs without Words,** Nos. 2; 4; 35; 41.

Greater independence of Part II is found in Mendelssohn, ** Song with-
out Words** No. 7;

And still more in Schumann, ** Album-Blatter ** op. 124, No. 8 (in which,
on the other hand, the outward connection between the Parts, i. e., at the
cadence, is closer than usual).

Grieg, Lyric Pieces, op. 12, No. 5 (Parts II and III repeated).

Chopin, Mazurkas 11 ; 24.

Examples of the 3-Part Song-form in which the diversity of character in
Part II is carried to an extreme, are rare ; and in no case, however rich they
may be in imaginative contents, or superior in attractiveness, can they be
regarded as models of logical structure, or as thoroughly genuine exponents
of the tripartite Idea^ as demonstrated in the foregoing paragraphs. See,
for instance, Beethoven, Bagatelle op. 33, No. 2, up to the "Trio**; and
Schumann, " Album-Blatter,** op. 124, No. 3. For the more reasonable
classification of such irregular forms, glance at par. 112a.

The introduction of new material into Part II is most easily accounted
for and justified when the latter is *' sectional'* in form; see Ex. 86, and its
notes.

Tonality of Part II.

88. As regards the modulatory design of the Second Part,
it should, in general, avoid the Tonic line of the original key; for

148

TONALITY OF PART II.

Par. 88.

the impression of. a Departure will depend as much upofn a change of
key as upon anything else ; and the idea of a Return is realized when
the original key is regained, after a more or less marked absence.

But the Dominant (or Subdominant) harmony of the original
key may be introduced into Part II ; and in some cases the Dominant
furnishes the basis of the entire Part from beginning to end ;
see Beethoven, Pfte. Son. op. 31, No. 3, "Trio" of the third
Movement; Symphony No. i, **Trio" of the third Movement.

Or the Tonic harmonies may be touched in passing into other
chords or keys.

Or the entire Second Part may disport itself in other related
(or even unrelated) keys. See Ex. 89 ; also

Mendelssohn, " Songs without Words " No. i (Part I, E major; Part II
begins in E minor, passes into G major, and remains there nearly to the end) ;
No. 7 (Part I, E[> major; Part II, E|7 minor, Q^ major, EI7 minor); No. 25
(Part I, G major; Part II, E minor, F major, G major transiently, A minor,
C major, A minor) ; Nos. 29; 31 ; 40; 42.

Beethoven, Symphony No. 2, **Trio" of 3rd Movement (Part II all on
the Dominant of the Relative key, up to the last chord).

For more specific directions in reference to Modulation, see par. 94.

As a rule, the Second Part should at least begin with some
other than the original Tonic harmony, in order to distinguish
itself immediately from the beginning of Part I. This need not
interfere with the plan of derivation and relation illustrated in
Ex. 81. When, as is the case in some rare examples, the Second
Part starts out exactly as the First Part began, the member should
diverge very soon into other chords or keys. For example :

Allegro,

85.

^

4r

*— »

^-^

*— »-

m

f=^=±

— I-

—f9-

^ — ^a

m

i

Par. 89a.

STRUCTURAL DESIGN OF PART II.

149

Part II.

r fJ7l || j r J

#*

^

*1)

Crete.

m

^m

* p f i TfTp ^fff*! f'if'f' ^ f'tff '

g

^

X X lJ

it

6 major C m^jor D m%jor E minor. . .

Jjjjjji

XXX

f

D lm%jor . . (Dominant of original key.) [

II Part m.

X X

Haydn.

*i) Part II is exactly identical with Part I, up to this ftj, which turns the
melodic current from G into C. The keys through which the Second Part
passes are indicated.

See also Schumann, op. 15, No. 7 (** Traumerei "); and Mendelssohn,
** Songs without Words,'' No. 36; No. 20.

Structural design of Part II.

89a.. In regard to the form and length of the Second Part,
it is both unnecessary and impracticable to recommend any other
law than that of sensible proportion, and balance ; and this, as it
will depend upon a variety of circumstances and conditions, must
be left to the judgment or intention of the composer. The principle
of approximate symmetry is apt to prevail, and therefore, as
already stated, if Part I is a Period, Part II is somewhat likely
to be a Period also, perhaps slightly extended.

Sometimes, however, the Second Part is so brief (only Phrase-
form) and comparatively insignificant, that it assumes the character

ISO

SECTIONAL FORM OF PART II.

Par. 89b.

of a mere interlude, — a, hasty (though genuine) Departure, intro-
duced for no other purpose than to create the impression of a Return.

See Mendelssohn, Fantasie op. 28, 2nd Movement, first 18 measures
(Part I, 7 measures; Part II, only 3 measures); op. 16, No. i. Allegro move-
ment (Part I, 34 measures and repeated ; Part II, only 10 measures).

Chopin, Mazurka 47 (pp. 68, No. 2) first 28 measures.

Beethoven, Violin-Concerto, op. 61, Finale (Rondo), first 18 measures
(Part II, only 2 measures).

Other examples of the small Second Part are cited in the footnote upon
page 138. (Beethoven, Pfte. Son. op. 49, No. 2, Finale^ first 20 measures ; etc.).

It is, perhaps, more common to make the Second Part a little
longer than Part I. And, unquestionably, the peculiar nature of
the Chain-phrase and the Group of Phrases^ especially in sequen-
tial succession^ adapts them in every sense more fully to the
purpose and conditions of a Second Part than any other more
regular and perfect structural design. Hence, the formation of
Part II in Ex. 85 may be regarded as exemplary. See par. 31, first
clauses, and the very last clause of par. 58.

See Mendelssohn, ** Songs without Words" No. 35 (Part I, Period;
Part II, Period extended) ; No. 28 (Part I, Period; Part II, Period, sequential
construction); No. 19 (Part I, small Double period; Part II, group of 6
Phrases); No. 31 (Part II, a Group of Phrases); No. 10; No. 15; No. 40
(sequential construction).

Sectional form of Part II.
(b) In some, especially the larger, varieties of the 3-Part
Song- form, the Second Part is sectional in form ; i. e., it consists
of two (or even more) "Sections," separated from each other by
a complete cadence in the momentary key, and often so distinctly
individualized that they would have to be regarded as complete
*' Parts " under ordinary circumstances ; but their location^ between
Part I and its Recurrence, renders it possible, notwithstanding
their independence, to preserve the impression that they are only
the subdivisions of one broad Part, — ^just as the three-fold condition
of Departure, Absence and Return may all be comprehended within
the single idea of Digression. For illustration :

Allegretto,

86.

Par. 89b.

SECTIONAL FORM OF PART II.

151

i

Part n. Section 1.

fe

r

S

#

)f.^i *

^
"

Crete.

f

P

y

^si=^ — ^•

X J; X -

f

-ftK#-

^^

^ X-

erMc.

m^

f

■* ^■

4—1* U I

1— <-

tt #

I

/

fr

1^

t=P

t=*:

Extension

in

^

li Section 2.

i

a a* L j.

^

3?e

//

-J^^^

j" -* vJjlJ ^

Extension

ii*"^T^tt

• -«^^*»» •

ki J I -J.

r^irr

• •

T^-

?

^

^

• •

-j^~2-

?z^

152

THE CADENCE OF THE SECOND PART.

Par. 90a.

Bbbthovbn.

fe^ j^j r-*-iJ

13L.

PP

%I

X X

^

w:x=t

I

Part in.

rallent.

etc.

1^-

i

f^

*i) Pfte. Son. op. 2, No. 2, Scherzo, — The subdivision of the Second Part
into two independent sections is strongly marked ; but they nevertheless-
constitute, together, only the one Part which represents the Digression, the
interim between Parts I and III, whose identity is unquestionable. Still, i£
carried to an undue extreme, this subdividing process (like that mentioned
in the very last clause of par. 87) destroys the legitimacy of the tripartite
design, and gives rise to a ** Group of Parts," such as appears in Chopin^
Mazurka No. 14, the classification of which will be found in pars. 114, 115.

See also Ex. 89; also Mendelssohn, op. 72, No. i.

Schubert, Pfte. Son. No. 10 (B(> major), Scherzo,

Haydn, Symphony No. 11 (Peters ed.), Menuetto,

The most plausible variety of this sectional form is obtained
by adding a Codetta to the body of the Second Part. This, while
comparatively unusual, and somewhat out of place, can nevertheless
be done without injuring the form, if the Codetta-section be chiefly
a confirmation, or independent extension, of the Cadence-member
of Part II, so manipulated as not to interfere with the purpose of
the latter to prepare for the Recurrence of the First Part (as Third
Part).

See Mendelssohn, "Song without Words'' No. 30 (Part I, Double
period, 15 measures ; Part II, Section i , Group of 4 — or 5 — Phrases, 20 measures ;
Section 2, Codetta of 4 measures; a 3rd Section follows, as Retransition, to
which explanatory reference will be made in the next paragraph).

See also Mozart, Pfte. Son. 17 (Cotta ed.) Finale^ first Subject (Part I^
Period, 12 measures ; Part II, Section I, Group of 3 — or 4 — Phrases, 18 measures ;
Section 2, Codetta of 2 measures, repeated; a 3rd section, — Retransition, —
follows, as in the preceding quotation).

Schubert, Moment musical^ op. 94, No. 6 (Part I, large Per., 16 measures ;
Part II, Per. ext. at end, 17 measures; Codetta of 3 meas., repeated; a Re-
transition of 14 meas. follows).

Beethoven, Pfte. Son. op. 13, Adagio^ meas. 23-27.

The Cadence of the Second Part.

90a.. As regards the cadence of Part II, it is evident that
it must be made in such a manner as to foreshadow, prepare for,
and lead with a certain degree of emphasis into, the first member

Par. 90a.

THE CADENCE OF THE SECOND PART.

153

of Part I as it recurs at the beginning of Part III. Hence, as
shown in par. 83a, 2nd clause, it is most likely to be made upon
the Dominant Harmony of the original key (or the Tonic Harmony
of the Dominant key) , because no other chord than this tends so
urgently toward the tonic Harmony, with which Part I (and
therefore Part III also) is supposed to begin. Sometimes the
Dominant idea (as Dom. chord, or Dom. key, or Dominant Organ-
point) runs through the entire Second Part, as in Ex. 78.

Beethoven, Pfte. Son. op. 31, No. 3, **Trio*' of the 3rd Movement;
Pfte. Son. op. 2, No. 2, Largo^ first Subject; Mendelssohn, **Song without
Words" No. 2.

Or the Second Part is conducted into the Dominant a certain dis-
tance (one or more measures, or Phrases,) before its end, as in Ex.83.

Often quite a persistent extension of the Dominant Harmony
is made, in some form or other, to stimulate expectation of Part
III, as in Ex. 84 and Ex. 85.

Chopin, Mazurka 18. Mendelssohn, ** Songs without Words," Nos. 9,
19. Beethoven, Bagatelle op. 33, No. i.

AUegretto.

Part 11.

87. <

I

^PP

^

t^

i mnrrrn

etc. ; regular
Period.

Sl^

Ef

I

m

f »-

■F — — 1—

p

t^

t t.

t .* } 1 ^

^ (ExpanAion of Dominant).

154

THE CADBNCB OF THE SECOND PART.

Par. 90b.

m

^

r-\-

Haydn.

#=P=

rm f T f r !=f=r

I

Partm.

etc.

dimin.

^

I

I

t

i

t

#— X— at-

r

f

Sometimes this expansion of the Dominant harmony is elabo-
rated by transient alternation with other (most effectively with
Subdominant) chords ; the Dominant is reached in due time, but
instead of passing immediately over into the Tonic, it sways a few
times, generally at least twice, back and forth, like a pendulum
before coming to rest.

See Mendelssohn, *' Songs without Words '* No. i, measures 9-14 from
the beginning of Part II ; No. 10, meas. 26-38 from beg. of Part II; No. 27 (the
expansion and elaboration of the Dom. differs so completely in character as to
appear to be an Interlude or Retransition).

The hearer's anticipation of Part III is furthermore confirmed
and emphasized, as a very general rule, by a relaxation of motion
{riiard.) in ending Part II. This is seen in Ex. 83 ; also in Ex.
86, where the effect is further heightened by a rest of nearly two
measures' duration, — the appropriateness of which, after so wide
a digression in character and in key, the student will realize.

Mendelssohn, " Songs without Words," Nos. 28; 32; 8.

On the other hand, the Dominant ending is sometimes so
brief, and the cadence so completely bridged over, that, but for
its being in the proper measure, and followed by a particularly
well-defined announcement of the original initial member, it could
not fulfil its important purpose.

See Ex. 76; Mendelssohn, ** Songs without Words," Nos. 23 ; 32 ; 38; 20
(very brief diminished 7th) ; Schumann, op. 15, No. i ; Chopin, Mazurkas,
Nos. 4; II.

(b) But, while the Dominant termination of Part II is unques-
tionably the simplest, the most common, and (ordinarily) the most
appropriate, other harmonic forms are quite frequently employed ;
chiefly because of their superior effectiveness, but also in consequence
of a possible irregularity in the beginning of Part I (and Part III)
which may call for special treatment of the cadence of the Second
Part.

Par. 90b. THE! CADENCE OF THE SECOND PART. 1 55

In Mendelssohn, ** Song without Words " No. 25, Part I begins upon
the Dominant, and consequently, the Second Part ends upon one of the
Second-Dominant* Harmonies (the altered IP); No. 46, though in G minor,
begins upon the Dom. of F; Part II closes with a very prolonged exposition
of this same chord, and leads into Part III very smoothly, though noticeably
enough ; No. 24 is similar, though more abrupt.

The most common substitute for the usual Dominant ending is
the chord (or key) upon the Mediant, — the 3rd scale-step of the
original key. This will appear less peculiar when the student
recalls that the chord III is a member of the Dominant harmonic
fatnily^ and that, in any event, it contains the leading-tone of its
key^ through which the safe transition back into the Tonic begin-
ning is sufficiently assured.**

See Mendelssohn, ** Song without Words '' No. 22 (F major ; the Second
Part ends upon the Tonic of A minor, which corresponds to the III of F ; the
e, in Bass, is effectively treated, as Leading-tone) ; No. 11 ; No. 42.

Chopin, Mazurka No. 16 (AI7 major; the Second Part closes with an
emphatic extension of the chord upon the 3rd scale-step, but in the major
form, as actual Dom. chord of F minor; this merely adds to the effectiveness
of the device, without impairing its legitimacy); Mazurka No. 29 (the same) ;
Mazurka No. 22 (Gjf minor; Part II ends upon the Tonic chord of its 3d step,
i. e., B, — but, naturally, as major Triad; this cancels the Leading-tone of the
coming Part, it is true, but the transition is still far less abrupt than in many
other cases) .

■

More abrupt forms of transition from the end of Part II into
Part III are found in Mendelssohn, **Song without Words"
No. 45 (hingjng upon the Leading-tone) ; No. 39 (Parts I and III
beginning upon the Subdominant Harmony).

Another effective mode of closing Part II, in convenient prepa-
ration for the initial member of the Third Part, consists in employing
some form of the Subdominant (or Second-Dominant) Harmony,
i. e., the IV, II, 11^ or IV^, either in their legitimate form, or
(preferably) in any of their altered or mixed forms (with lowered
6th scale-step, raised 4th scale-step, etc.). These chords almost
necessarily involve resolution into some inverted form of the Tonic
Harmony, wherefore Part III will, in this case, usually be found to
begin upon the Tonic six four chord. For example :

*See the Author's ** Material used in Mus. Comp.*' paragraphs 6c,
** /(Um, paragraphs 90-93.

156

THE CADENCE OF THE SECOND PART.

Par. 90b.

A ndante.

Partn.

Mbndblssohn. No. 48.

Part III. , t^r*— ^ ^

'iff r'r ' ^^

i H I y

fla *1) I,

*i) The Second Part closes upon the chord of the IP (with the raised 2nd
and 4th, and then the lowered 6th, scale-steps); and Part III begins upon the
•accented \ chord of the Tonic. This latter differs slightly from the harmonic
form of the beginning, but not enough to obscure the effect of Recurrence.

See also Mendelssohn, ** Song without Words " No. 7 (Part II ends with
an altered Subdom. chord; Part III begins with the Tonic J chord); No. 12;
No. 29; No. 43 (all similar).

Any such tampering with the beginning of Part I (as Part
III) involves, of course, the danger of impairing or destroying the
principal landmark of the tripartite form, i. e., the distinctly
recognizable Return to the beginnings in exactly its original condi-
tion. Therefore, if any modification at this point be ventured, it
must be very cautiously effected, with a view to preserving at least
all the essential and characteristic elements of the initial member of
Part I. To what such modification of the latter may lead, will
be seen in par. no.

Par. 90c. THE RE-TRANSITION. 157

See also Mendelssohn, *' Songs without Words," Nos. 41 ; 40; 37 ; 21 ; in
each case the initial melodic member is retained in Part III, but differently
harmonized. No. 36 (the first measure of Part III is so disguised that full
consciousness of the form is not aroused until the 2nd measure appears).
Chopin, Mazurka 46, Principal Song; Nocturne No. 16, first 12 measures
(already referred to as 3-Part Period) .

The most irregular and confusing of all modes of terminating
the Second Part, is to lead it into the Tonic Harmony ; this, just
before the announcement of Part III upon the same chord (Tonic),
seriously endangers the identity of the Third Part as * * Return ' '
(comp. par. 88, first clause).

See Ex.91 ; and Mendelssohn, '* Songs without Words," No. 4; No. 31 ;
Beethoven, Rondo in C, op. 51, No. i, first Theme; (Rondo in G, op. 51,
No. 2, first Theme, is scarcely better).

The Rk-transition.

(c) It must be borne in mind that a full-fledged Second Part
comprehends three successive phases, which, while their lines of
demarcation may be almost imperceptible, are still distinct enough
in their several purposes to give rise to three different courses of
conceptive action. These are : (i) a Departure from the line of the
principal Part; (2) an optional period of Absence, during which
considerable individuality may be developed, — as, for instance, in
the sectional form ; and (3) the Return to the original starting-point.
This third phase may justly be regarded as the most significant of
all, for it is the ultimate aim of the entire Part, and an endless
amount of ingenuity may be displayed in the effective formulation
of the means employed in realizing this ' * return to the starting-
point." The larger the form, the greater the likelihood and
necessity of devoting a separate section wholly to this purpose.
For such a * ' returning section ' ' the term Rk-transition appears
most fitting.

In case the Re-transition thus forms a separate section, by
itself, the body of Part II will usually have a complete Tonic
cadence in whatever key it chances to have reached. The Re-
transition then leads back, from that point, into the key, chord,
style and melody of the beginning, in such a manner, and at such
length, as seems most adequate and effective. It appears useless to
undertake to give any more definite rules than this for the process ;
for it is governed by ever- varying circumstances. There is no

158

THE RK-TRANSITION.

Par. QOc

better source of information than standard musical literature, and
the student is therefore urged to make careful examination and
analysis of the given examples, references, and whatever other
illustrations he can find. In all essential respects the Re-transition
partakes of the nature and purpose of the ** Introduction," though
infinitely richer in possibilities. For example :

AlUgro.

m

88. <

pK^

-i— IK-

^yrH-3— y
^= ^^1^

-9^

\ — V

J, ' J.^ ^ ^

^

n-

-9 r

jr—t.

*— -5

J=F^

*=t(^

5

:5a?

wft

jrr

Y».

K

i

-«-

3

^^- -

-a=^

1—1-

^^b it t ir^

i^ii

*=*=*=:*

^^^S^'^

-#— Im-

part n. Section 1.

m

V

pf f f ^ I* te

t^

*i)

»

^

^

-0—^-

— ^-^

1— r

ir?=y

Par. 90c.

THE RE-TRANSITION.

159

Section 2.

Cad. II Betranaition.

.Part m.

Schubert.

*i) The division of this Second Part into Sections is very recognizable.
Section i is the ** Departure," with just sufficient lingering recollection of
Part I to emphasize the idea ; Section 2 is a more decisive digression from
Part I, and stands for the *' Absence."

*2) Here the body of Part II terminates, with the perf. cadence in the
key into which the 2nd section has strayed (G\} major). The Re-transition
which follows, consists: firstly, in an abrupt chromatic dash back into the
original Dominant, accentuated by the isolated forte; and, secondly, in
a 4-measure expansion of this Dominant, during which the Soprano and Bass
converge smoothly but persistently toward the tones with which Part I began.

See also Schubert, Pfte. Son. No. 4, Menuctto.

Mendelssohn, **Song without Words" No. 13, E|;> major; the body of
Part II terminates in the 32nd measure, on the Tonic of G minor ; the following
8 measures are Re- transition, based entirely upon the first melodic figure of
Part I, which suggests, and is skilfully led into, the Third Part; the original
Dominant is reached in the 5th measure of the Re-transition, and then sways,
as seen in former references,-^text following Ex. 87 ; — it will he observed that
wheUy in thus swaying^ the original Tonic is touched, the composer is careful

l6o THE RE-TKANSITION. Par. 90c.

to use the minor form of Tonic Harmony^ in order to preserve, and even
increase, the force of the original major form, when the latter heralds the
beginning of the Third Part; this same significant trait will be found in many
other major examples, — *' Songs without Words," Nos. i, 7, 19, 20, 31, 37.

See also *' Songs without Words," No. 16; Part II virtually ends in the
19th measure, with a Dominant semicadence ; a Re-transition of 2 measures
follows, built upon the first melodic member. — No. 26; effective Re-trans, of
4 measures before Part III. — No. 34, measures 25-29. — No. 37, Re-transition
of 4 measures. — No. 3; Part II is 16 measures long, ending with a firm
cadence in the kej of the original 3rd step; a Re-transition of 5 measures
follows.

Such a separate Re-transition is very likely indeed to accom-
pany (i. e., follow) a Codetta to the Second Part, when that
somewhat rare factor is employed (see last clause of par. 89). The
Re-transition, in this case, starts with the apparent intent of
repeating the Codetta^ but sooner or later "dissolves" the original
form of the latter, and digresses into such a course as defines its
re-transitional purpose.

See Mendelssohn, "Song without Words" No. 30, A major; Part II
terminates with a Tonic perf. cad. in E major in the 35th measure; the
4-measure Section which follows is an independent and regular Phrase, with
its own unmistakable cadence ; it does not belong to the body of the Second
Part^ for that terminates with the utmost decision before this Phrase begins ;
nor is it in the Third Party for obvious reasons ; nor is it a Re-transition into
the latter, because of its complete cadence ; thus it proves to be an adjunct of
Part II, in the capacity of a " Codetta." The measures which follow indicate
for a time the intention of repetition, but then abandon this purpose and
become a genuine, and exceedingly attractive and clever Re-transition.

The same treatment is found in ** Song without Words " No. 17, measures
14-17 (Part I closes in the 4th measure with a vague form of cadence, ex-
plained in par. 93rt).

See also Chopin, Nocturne No. 17, measures 21-28.

In the ordinary 3-Part Song- form, however, the Re-transition
is somewhat more likely not to constitute such a distinct section,
but to grow out of some later member of the Second Part, perhaps
so imperceptibly that it is not easy to point out the exact beat
where the purpose of re-transition is evident. Here, the persistent
bent of the harmony toward the original Dominant^ and the equally
characteristic tendency of the melodic current to lead into and
regain Xh^ first tone and figure of the First Part, is possibly more
noticeable than in the separate Re-transition ; and^ in any case^ this
two-fold tendency (first toward the key and then toward the melody^

Par. 90c.

THE RE-TRANSITION.

l6l

must be regarded as inherent in all re-transitional passages. This
is plainly shown in Ex. 89. In the following illustration, the
bent of the Re-transition is concentrated upon the initial melodic
member of Part III :

A ndante.

n n r =^

3

Bb 1

Part II

etc. See Ex. 63, or the Original. *1)

V

gf^^

Mendelssohn.

*i) ** Song without Words '* No. 33.

*2) The re-transitional purpose appears to assert itself here, but in such
close keeping with the melodic line of the foregoing measure, that its actual
bearing upon the coming Third Part is only gradually apprehended. It is left
to the pupil to trace, in this particularly beautiful example, the slow but sure
approach of the melodic delineation to that of the first figure of Part III, and
to observe how the harmonic color of these re-transitional measures tinges
the first tones of the Recurrence of the initial figure, until the Cft puts a stop
to the playful illusion, and restores the serious mood of the Third Part.
** Song without Words '' No. 41 is similar.

SeealsoMENDELSSOHN,**Songs without Words," No. 11 (there-transitional
purpose is manifested in the 20th measure ; the melodic figure of three tones
corresponds to the beginning of Part III); No. 14, last 3 measures of the
Second Part; No. 21, last 8 measures of the Second Part (built upon the first
melodic member, — 2 measures, — of Part III) ; No. 39, last 4 measures of the
Second Part; No. 45, C major, last 4 (or perhaps 8) measures of Part II;
No. 37 ; No. 36, last 3 measures of Part II (this re-transitional expansion of the
Dominant flows over into, and slightly influences, the first figure of the Third
Part). No. 47 (A major) is peculiar; Part II virtually ends at the beginning of
the 37th measure ; the measures which follow are partly a sequence of the two
which precede, and partly an anticipation of the beginning of Part III.

1 62

THE THIRD PART.

Par. 01.

The Third Part.

Q I , In the ordinary complete 3-Part Song-form, here under
consideration, the Third Part is a simple recurrence or

REPRODUCTION OF PaRT I.

If the First Part closes with a perfect cadence in the original
key, then it may reappear literally in Part Third, as conventional
'* da capo,"

But if the First Part cadences in some other key, or has any
other imperfect form of cadence, then at least enough of the ending
of Part III must differ from the original Statement to admit of its
cadencing properly, in the principal key.

And, furthermore, any purely unessential modifications of
Part I (affecting the harmonization, or the register, or ornamenting
the Melody) may be introduced in the Third Part, to heighten its
effect. But no more important changes than these are exhibited
in the " Simple Complete Form." For example :

Poco sosttnuto.

81. <

V=t

I I I— I— I.

ztr=t=t

U Lfi?|Ld"{?^^=ff^ ^^^

Part n.

Par. 91.

THE THIRD PART.

163

^* t u

0$

<

m

I

^

PartllL

DvoftXK. "AVtt» IVorW Symph.

*i) Part II closes on the original Tonic ; but the beginning of Part III
is distinct enough. See par. 93^.

*2) Part III is essentially the same as Part I ; the register of the melody is
changed, and one measure is omitted. — In Chopin, Mazurka No. 49, Part III
is not written out at all, but simply indicated by the words *' da capo."

In Ex. 76 and Ex. 77, though a smaller species of the tripartite design, the
literal agreement of Part (Phrase) III with Part (Phrase) I, is illustrated.
In Mendelssohn op. 72, No. i, and ** Song without Words" No. 45 (C
major) the Recurrence is literal. — ^Ex. 89 is concluded (in the Original) by

164 EXTRANEOUS MEMBERS. Par. 92.

reproducing the first 4 measures of the melody an octave higher than at first,
while all the rest is left precisely as it was. — Ex. 84 is concluded in the Original
(Schumann, op. 68, No. 24) by reproducing Part I with very slight changes. —
In the conclusion of Ex. 83 (Chopin, Nocturne 15, Principal Theme) a brief
variation of Part I is made.

In Beethoven, Pfte. Son. op. 2, No. 2, Scherzo (Ex. 86) the reproduction
is literal. In Pfte. Son. op. 22, Menuetto, slight changes are made. In the
F major Variations, op. 34, Theme, the Recurrence is literal.

In Mendelssohn, ** Song without Words" No. 22, Part III is a little
differently harmonized from its First Part; in No. 32, the single melody of
Part I is transformed into a duet in Part III.

In Haydn, Pfte. Son. No. 6(Cotta ed.), ist Movement, Principal Theme,
there is, in addition to interesting variation in Part III, a change in the last
member, necessitated by the circumstance of Part I having closed with a
cadence in the Dominant key; the same in Schubert, Pfte. Son. No. i, 2nd
Movement, Theme (Part I repeated) ; the same in Beethoven, Pfte. Son.
op. 2, No. I, ** Trio " of 3rd Movement.

Chopin, Etude in A[> (No. 3 in the ** Moscheles-F6tis " method). Part III
slightly expanded at end. Chopin, Mazurkas 16; 22 (long Second Part); 24
(Part II quite distinct, and repeated); 40; 43 (Part II repeated, distinct
re- transitional interlude; Part III, literal Recurrence of Part I); 44.

Grieg, Lyric Pieces op. 12, No. 2; op. 38, No. 2; op. 43, Nos. i, 4,6
(extended at end) .

Brahms, op. 116, No. 7, "Trio" (6-8 Time). — Op. 117, No. i, first tempo
(Parts I and II small Periods; Part III is extended to Double period, by
reproduction; Codetta); compare with 3rd (last) tempo. — Op. 118, No. 2,
meas. 1-25 after first double-bar. — Op. 118, No. 3, first 37 measures. — Op. 118,
No. 6. — Op. 119, No. I. — Intermezzo, op. 76, No. 4. '

Extraneous Members.

92. The 3-Part Song-form may be, and, as a rule, should be,
enlarged by the addition of a Codetta or Coda. More explicit
explanation of the character, purpose and formation of these factors
is given in par. 98, to which brief reference may be made.

The prefixing of an Introduction, while less significant than
a Coda, is, however, barely less desirable ; for it serves to empha-
size, by mild contrast, the body of the Song-form, to which it stands
in the same relation as the *' margin " to a picture.

Turn, again, to Mendelssohn, **Song without Words" No. 12, and
examine into the relation of the Introduction to the entire composition ; also
Nos. 15; 26; 29; and 32.

When a Prelude is substituted for the Introduction, or a
PosTLUDE for the Codetta, its architectural and logical connection
with the body of the form is less close, and its relation to the latter
is more like that of the independent " frame " to the picture.

%

0^

Par. 02. EXTRANEOUS MEMBERS. 165

Examine, again, Mendelssohn, ** Song without Words** No. 4 (both
Prelude and Postlude ; almost identical ; here the simile of the picture with
its frame is very fitting) ; also No. 9 (a Codetta of 4 measures is added to
Part III, or, more accurately, to the body of the form, before the Postlude
appears); Nos. 16; 28; 35; and 41 (in each, similar Prelude and Postlude).
In No. 21 the Prelude does not recur at the end, and, in other respects also, it
approaches the character of an Introduction. — In No. 3 the Prelude becomes
the thematic basis of the entire elaborate Coda.

The insertion of Interludes between the Parts is a more

hazardous proceeding, as they tend to disrupt the structural con-

inuity. But if carefully handled, with a view to avoiding this

error, they may be very effective. They must either conform

strictly to the limitations imposed upon the Introduction (in par.

28, second and third clauses) ; or they must prove their extraneous

nature by agreement with a Prelude.

In Mendelssohn, ** Song without Words** No. 23, the substance of the
Prelude reappears after 9 measures, as Interlude between Parts I and II ; then
again 23 measures later, as Interlude between the second Division of the form
(i. e.. Parts II and III) and its repetition; and again as Postlude, at the end. —
In No. 27 a similar recurrence of the substance of the Prelude appears between
Parts II and III (quasi as Codetta or Re-transition), and again at the end. —
In No. 29 there is one measure of Interlude between Part I and its repetition ;
in No. 46 the same (in each case borrowed from the Introduction). In No. 34
the brief introductory figure is utilized as Re- transition, both in the second
Division (Parts II and III) and its reproduction ; and again, as final section of
the Coda.

EXERCISE 23.

A, Take one of the major Periods with perf. cad. in the original key,
invented in Exercise 9; to this, as Part I, add a Second Part according to the
directions given in par. 87^, in Period-form, with cadence as prescribed in
par. 90a; and add a Third Part as literal Recurrence of Part I (the words
** da capo ** will suffice).

B, Some former minor Period with regular perf. cadence, as Part I;
to this add a Second Part according to the directions in par. 87*, in Period-
form ; cadence as in par. 9o« ; Part III a literal da capo.

C. Some other former major Period as Part I ; to this add a Second Part
according to par. 87c, in Period-form with extended Dominant cadence. Part
III a literal da capo,

D, Some other former (or new) minor Period as Part I ; to this add
a Second Part according to par. 87^/, in Phrase-group design, with some form
of cadence explained in par. 90^ ; Part III a nearly literal da capo. Review
par. 88.

l66 IRREGULAR CADBNCBS.

EXERCISE 24.

Par. 93a.

A, Some former (or new) Period in major, with cadence in some other
than the orig. key, as Part I ; to this add a Second Part according to par. 87^,
in Phrase-group design, with optional cadence-form; Part III a literal da
capOy excepting change of last member. Review par. 91 .

B, Part I as minor Period (regular or extended), with cad. in some
related key; Part II optional melodic construction, but sectional form, as
explained in par. 89^; brief Re-transition, par. 90c; Part III a slightly
modified Recurrence of Part I. Review par. 88; and the directions given
IN Exercise 9.

C. Part I major, optional form, but with Introduction; Part II optional
form and structure, but with Retransition ; Part III a slightly modified Recur-
rence, followed by ^r«^ Codetta. Review par. 92.

D. Part I minor. Double period. Introduction or Prelude; Part III
a slightly modified Recurrence with brief Codetta or Postlude.

EXERCISE 25.

A, Part I major, optional form, but with brief Codetta; Part II optional ;
Part III a modified Recurrence, including former Codetta, and with an
additional brief Codetta. Review par. 23 ; 27 ; par. 30 ; 53 ; par. 40.

B. Minor, entirely optional in all save the essential requirements of the
Simple Complete 3-Part Song-form.

t

CHAPTER XIII.

ADDITIONAL DETAILS OF THE SONG-FORMS.

I. Irregular Cadences.

93. There are a few peculiar conditions of the cadence, not
yet explained, to which attention must now be drawn.

(a) The purpose of a Cadence is, to indicate a point of more
or less complete separation between two contiguous members of
the structure, — a *' joint," where one member ends and another
begins (par. 2b). There are two modes of fulfilling this purpose,
namely :

(i) to mark the end ofthejirst of these two members distinctly;

Pai.Ma. IRREGUI-AR CADSNCSS. 1 67

(2) to mark the beginning of the new member with equal dis-
tinctness.

The first mode is the simpler and more common, and consists
in using a strong Cadence-chord, and making a pause in the audible
rhythmic pulse, adequate to the degree of separation required.
This is illustrated in nearly all of the foregoing examples; see
Ex. 46, No. 3, and the end of any example from 25 to 37.

The second mode, while naturally less distinct and decisive
at the moment, is nevertheless, when properly executed, ultimately
quite as effectual, and, at the same time, smoother than the other.
It consists in defining the beginning, at least, of the following
member (Phrase, Period, Part, or whatever it may be) so distinctly
that no doubt of its being a new member can exist ; this defines
the form, for where there is a beginning there must before have
been an ending. This is precisely the principle which prevails in
the mode of ** phrasing" explained in par. 10, 3rdly (Ex. 15);
while sometimes desirable even between ''Parts," it is naturally
most important in application to the arrangement of smaller mem-
bers, hec2M^e frequent complete stops (to define the **end" of
every member) would entirely disrupt the design.

The more unmistakable the outset of the following member is, the less
emphatic need the foregoing Cadence be. This accounts for all the vague,
imperfect Cadences (often so brief and indefinite as to be scarcely discoverable
at first as Cadences) hitherto seen, and those now to be examined.

For example, the Semi-cadence at the end of Part First in the Song-
forms: See Mendelssohn, ** Song without Words" No. 17, measure 4 (reg.
Dom. Semicadence; but evidently the end of Part I, because what follows
is sufficiently characteristic to be evidently the beginning of Part II) ; No.
38, measure 10 (Cadence of Part I on a | chord ! It is rendered sufficient by
appearing in the 8th measure and being repeated in the loth ; moreover, the
beginning of Part II is unmistakable); No. 47, measure 18, similar to No. 38;
No. 39, Part I ends with chord-Fifth in Soprano. In Nos. 7, 8, 10, and 11,
Part I ends with some imperfect form of Cadence (usually extended y however);
but the identity of the Part is fixed by its repetition (see par. 77, last clause). In
Nos. 30, 33, 41, and 48, the Cad. of Part I is unusually brief and ** breathless,"
but still sufficient, under the thematic circumstances. .These stand in marked
contrast to the decision and frankness of the Cadence to Part I in Nos. 23, 12,
15, 27, 35 and others.

For other examples of an imperfect Cadence at the end of a complete
member, see Mendelssohn, No. 15 ; at the beginning of measure 5 from the
end, the Tonic Cadence-chord has the Third in Soprano; but the following
member nevertheless constitutes a Codetta (without being preceded by a
compl. perf. Cad. according to rule), because its outset is sufficiently distinct

1 68

IRRKGUI.AR CADENCES.

Par. 93b.

to indicate the beginning of a new member. The same trait is observable in
No. i6, in the 4th meas. from the end; No. 31, five meas. from the end; and
No. 33, six meas. from the end.

See also Chopin, Nocturne No. 9 (op. 32, No. i), ending (the Codetta
follows an Evasion of the Cadence); Bkethovkn, Pfte. Son. op. 31, No. 3,
Menuetto (Part I ending with Dom. Semicadence, but repeated) ; Beethoven^
Pfte. Son. op. 27, No. i, Adagio (Part I ending with a Semicadence); Haydn,
Symphony No. i (Peters ed.), Menuetto^ Part I; Brahms, Symphony No. i^
AllegrettOy first 18 measures (probably a Two- Part Song-form, followed by
a somewhat extended complete repetition; Part I ends in meas. 10 with
a Semicadence); Beethoven, Theme of 24 Variations in D major, Part I;
Schubert, Pfte. Son. No. 7, Allegretto^ Part I ; Schubert, Moment musicaly
op. 94, No. 5, Part I; Chopin, Preludes op. 28, No. 11 and No. 22, Part I of
each; Mendelssohn, Praeludium op. 104, No. 2, Part I.

To this category belong also all examples of the " Elision '* (par. 60, which
review), in which the "Cadence" disappears altogether by the overlapping
of the contiguous members. See Mendelssohn, Prelude op. 35, No, d
(Part II begins in the middle of measure 15, with an Elision of the expected
Cadence); also the Praeludium of the ** Praeludium and Fugue in E minor**"
(without opus-number), measures 22-23.

(b) Exactly the reverse of this peculiarity (that just explained
being : a vague Cadence sufficing to separate independent members)
is seen in those places where a complete Tonic Cadence fails to sever
the connection between contiguous members, because the second one
maintains so intimate a thematic relation that it necessarily still
belongs to the section in question. Under such conditions it is
possible for a regular perfect Cadence to occur in the course of the
J^irst Part (i. e. without terminating it). As contradictory of
the strict rule given in par. 70 as this appears to be, it is not at all
unusual, — especially near the beginnings where such an unlooked-for
Cadence seems rather to serve the desirable purpose of establishing
the tonality. For example :

X. Moderato.

Par. gsb.

IRREGULAR CADENCES.

169

<

(Bi

^^m

:^

^

f

A i A A

fe

J i- ^

i

J-

1

<

i

Part II.

L I I

^

Folk-song. a. AlU^^retU.

^

rrrr

i

etc.

I

i

J^

«=

i

r T \

i

:5F=W

LriiKrtir ; ^;

g

<»-*? V-

■^m

Mozart.

*i) This is a Tonic perf. Cad. of the most unequivocal kind ; and yet it
does not terminate the First Part, because the latter obviously runs on, in
unbroken thematic connection, to the 8th measure. See also Ex. 62, 4th
measure .

*2) In this unique example the Part begins with a genuine perf. Cadence !
See the Original (Symphony in C, ** Trio" of Minuet).

See also Mendelssohn, "Songs without Words" No. 39 (perf. Cad,
in 2nd measure) ; No. 43 (Part I, extended Period, with perf. Cad. at end of
Antecedent Phrase); No. 44 (Part I, the same; in the last two cases the
unbroken thematic connection is established by parallel construction); No. 42
(Part I, contrasting Double period, with Dominant perf. Cad. in the center).
A glance at the beginning of the other ** Songs without Words," on the other
hand, teaches how, as a rule, the composer aims to avoid this misleading use
of an untimely perf. Cadence.

See further, Chopin, Mazurka No. 24, Part I ; Schubert, Moment musical
op. 94, No. 4, *'Trio" (5-flat signature) first 12 meas. (group of 3 Phrases,
repeated) ; Beethoven, Pfte. Son. op. 31, No. i, Adagio (Part I a very broad
Double period, parallel^ with Tonic perf. Cad. in the center); Pfte. Son. op.
27, No. 2, ist Movement, Part I (23 measures in length — including a 5-meas.
Prelude, — and containing tivo perfect Cadences in its course).

170 MODULATION. Par. 94a.

This trait is very common in repetitions. See Mendelssohn, ** Song
without Words" No. 4 (the perf. Cad. in meas. 4 does not mark the end of the
First Part, because the 4 following measures are a repetition); No. 11 (at
the beginning of the 5th meas. from the end there is a strong perf. Cadence,
but the following member is not a Codetta, because it belongs to the foregoing
Phrase, as evident repetition of the 2nd half).

(c) It is clear that both of the above irregularities of Cadence
are counteracted by the thematic conditions attending them. In
each case it is the unmistakable import of the thematic arrangement
(in reference to melody, rhythm, character, etc.) that overpowers
the ordinary forcible agency of the Cadence, and renders it at one
time unable to prevent, and again unable to effectuate, the *' sepa-
ration " of members, which is its chief purpose in form. This
evident predominance of the Idea over the technical Detail is an
additional corroboration of the justice of the classification defended
in the footnote on p. 138.

2. Modulation.

94. In addition to what was said in par. 88 about the modu-
latory design of the Second Part, the student must be reminded
that changes of key are very important and necessary indeed, in
all of the Parts (least, it is true, in Part I). This refers not only
to complete, or lengthy, modulations, but particularly to transient
digressions, which heighten the color of the Phrases, without
cancelling the ruling effect of the principal, or dominating, tonality.

(a) These modulations (i. e., transient ones,) need not even
be limited to the five Next-related keys,* but may extend to any
of the Remotely-related ones,** at least occasionally, and, of
course, on condition of returning immediately (or soon) to the
original tonality, or to its close neighborhood. The Remotely-
related keys are : (i) the Opposite mode of the same tonic (for
instance, C major-C minor, and vice versa) ; (2) the Stride (C
major— f minor, and vice versa) ; (3) The Mediant-keys C— E, C— Ab ;
C-Eb, C-A, etc.) ; and to these may be added, (4) all keys
accessible through the coincidence of important pivotal tones.

A superb example of this quality of modulation will be found in Schubert,
Impromptu op. 90, No. 3 (G major) measures 26 to 10 from the end (first
Phrase of Coda, and its repetition) ; the keys touched in close succession are
G maj.-c minor-G maj.-c min.-ab minor-G.

* •* The Material used in Mus. Composition,'* paragr. 266.
**The same, paragr. 267-269; par. 281, 282, 288; par. 289.

Par. 94c. MODULATION. 1 7 1

(b) In making more lengthy modulations, care must be taken
not to remain in any other key (whether related to the original key
or not) long enough to create the impression of a central or primary
tonality at that place. The original tonality, — the principal key, in
which the composition is written, — must be regarded as the axis,
about which the entire structure revolves. Any protracted absence
from this axis, or, more properly, any undue retention of some
o^her key, would have the effect of shifting the axis, or of replacing
it by a foreign one, and would thus tend to destroy the necessary
centralization of the keys and chords.

On the other hand, a temporary shifting of the modulatory
axis (subject to the above restriction) may constitute one of the
best modes^ of individualizing certain Divisions of the architectural
design, — for example, the Second or Middle Part in the tripartite
forms. Hence, some other key (probably related to the principal
one) may be chosen upon entering a new Part, or still larger
Division, and be allowed to prevail as temporary axis of that
section ; and the extent of the new key, and its distance from the
original key, will serve to determine the degree of independence
and individuality of the section in question. See par. 88.

This is illustrated in Schumann, Humoreske, op. 20, first Movement
(3-Part Song-form with Coda; Part I, repeated Phrase in the key of "two
flats'* ; Part II, a Group of three Phrases, throughout which the key of ** six
flats" prevails; Part III, literal Recurrence of Part I).

(c) Two very useful general rules of modulation in **Form"
are :

(i) Avoid a series (especially a long series) of modulations
in the same direction (i. e. successively increasing sharps, or
flats ; or successively decreasing the same, in more or less regular
progression) ; modulate beyond the desired key, if possible
quickly, by some extraneous transition, and then turn back into
the latter. For instance : if aiming for four sharps, from C
major, modulate into jive sharps, and then turn back to four.*

(2) After any remote modulation, it is better to turn hack
into some intermediate key, than to remain upon the new key, or
to pass on beyond it.

*It is well for the student to bear in mind that the fundamental modulatory
movements are best defined by signature, not by key-note.

i*

172 THE DYNAMIC DESIGN. Par. 95a.

Both of these must be accepted as general rules onlj, subject to many-
exceptions. As usual, the best source of information for the student on this
point is classic or standard musical literature, to the analysis of which he
should devote himself most assiduously.

(d) The general modulatory design of an entire composition,
almost regardless of its size, is as follows :

At the beginning, a sufficient pause in the original key to
establish the principal tonality ; then a general tendency upward
(toward the Dominant keys : those whose signatures contain one
sharp more or one flat less than the original signature) , ^-extending
approximately one-quarter or one-third of the entire distance ;
then a more or less ample digression, or a modulatory oscillation
between various keys in optional direction and measure ; and
finally, somewhere during the last sections, an inclination toward
the Suhdominant keys (those whose signatures contain one sharp
less or one flat more than the original). See par. 73c, 2nd clause.

This plan is quite. accurately followed in Haydn, Symphony No. 3 (Peters
ed.), Finale y first 20 measures.

See also Mendelssohn, ** Song without Words *' No. 48 (Part I, C major,
passing into "one sharp*'; Part III, "natural scale"; Coda, largely "one
flat'*); also Ex. 28(4 measures) where it is exhibited on a very small scale;
Ex. 47, No. I ; Ex. 71; 72; 75. There is scarcely an example to be found in
classic music, in which this general modulatory line is not more or less com-
pletely and emphatically traced, not only in the whole, but in each independent
;: Division.

» For examples of peculiar modulatory design, see Mendelssohn, " Wed-

ding-March ** (in C ; Part I begins with a distinct exposition of e minor) ;
J Beethoven, Pfte. Son. op. 26, Scherzo (A(7 ; Part I begins with exposition

of E|;>) ; op. 27, No. 2, Allegretto (the same idea) ; Pfte. Son. op. 90, 2nd
f. Movement (unusual prevalence of principal key throughout entire Principal

5, Theme); Violin-Concerto, op. 61, 2nd Movement (the same idea); Symphony

' No. 3, 3rd Movement (strangely meagre representation of the principal key, up

II to the jTin the course of Part III). Ex. 91 contains no modulation whatever.

\

3. The Dynamic Design.

95. The dynamic design refers to the use and disposition of
[! the various grades of tone-power, i. e., pianissimo^ piano^ forte^

S fortissimo^ with all their attendant shades of force, and all the

auxiliary tone-effects to which these distinctions may give rise
{crescendo^ diminuendo^ sforzando^ etc.).

(a) While it is quite certain that no definite rules for
a dynamic design can be, or should he^ given, it is equally certain

Par. 96. CONTRAST. 1 73

that some such design nevertheless exists in every good composi-
tion, whether defined beforehand, or in the course of conceptive
-action; and it is necessary for the student to realize and correctly
estimate the importance of these factors as a means of adding
life to the musical image, and of emphasizing essential traits of
the architectural design, — ^just as the effects of light and shade
add vitality and distinctness of character to a picture, a landscape,
or a theatrical scene. And, furthermore, there is an element of
suggestion in the definite purpose of making a crescendo or dimi-
nuendo^ di pp or^, which acts as a stimulus upon the imagination,
much more powerfully than might be suspected before putting it
to the test.

The general effect of these dynamic distinctions is fully enough understood.
Attention is simply called to the change in power as a means of increasing
the interest in a repetition (as seen in Ex. 56) ; and as a means of accentuating
the beginning of a new section by a sudden change from/ tofovff^ or from
yto /; see Beethoven, Pfte. Son. op. 2, No. 3, first Movement, measure 13;
also measures 25-27; Ex. 65, note *i); Ex. 95, note *i); Ex. 72; Ex. 73; Ex.
75 ; Ex. 88, beg. of Part III ; Ex. 89, note *2). — Observe, also, the dynamic
traits of Exs. 64, 74, 76, 83, 86, 91 and 99; and of all future quotations.

(b) Hand in hand with these factors, often almost inseparably,
go those shades of expression which refer to the rate of motion,
(i. e,^ritardando^ accelerando^ sostenuto^ tneno or piii mosso^ etc.).

See Ex. 83; Ex. 86; the rs\ in Ex. 56; Ex. 76; Mendelssohn, "Song
without Words" No. 39, last 11 measures. Compare par. 97^.

4. Contrast.

96. In paragr. 40, which review, the significance of Unity
and Variety was demonstrated, in their bearing upon the smaller
factors of creative structure.

As the dimensions increase, these same elements are more apt
to be recognized as the attributes of symmetry and contrast ;
and while variety and unity are originally of equal importance,
they gradually and steadily diverge, so that in proportion to the
magnitude of the design^ the necessity for Sytnmetry decreases^
and the need of the element of Contrast becomes more and more
imperative.

It is so obvious that Symmetry is of less importance in musical form than
Contrast, that some composers deny its claims altogether, and labor to avoid
all evidences of symmetrical structure in their works. But any such extreme

1 74 STYLE. Par. 97a.

conception must be condemned. It is more just to moderate this aim by
merely avoiding the obtrusive evidences of Symmetry ; and, at the same time,
to restrain all exaggerations of Contrast by constant and serious regard for the
vital condition of Unity.

This law is clearly recognizable in all the given details of
the Part-forms, whence, for instance, the inference may be drawn
that it is not nearly as necessary to preserve equality of size in
successive Parts ^ as it is in successive Measures; and that
contrast in style and character will be more pronounced between
consecutive Parts than between consecutive Phrases. Hence it
is that the student should now begin to devote a part of his
attention to the elements of variety and contrast ; and while he
must scrupulously observe moderation, Unity, and a rational
degree of formative and logical agreement, he must not forget that

THE BANE OF ALL ART-CREATION IS MONOTONY.

Glance at the evidences of Contrast in Ex. 48, No. 2; Ex. 53; Ex. 60
(Bkethovkn, Pfte. Son. op. 2, No. i, Finale^ ist Theme); Ex. 83; Ex. 86. —
Observe, in Mendelssohn, ** Song without Words" No. i, how welcome the
change in the 5th measure of Part II, ip a composition where uniformity
of style becomes almost oppressive. Note the brilliant effect of the Coda in
No. 3, with its change in rhythmic style. Also the interludes in No. 23. —
But observe, too, how wisely the unity of the whole is guarded, in each case.

5. Style.

9T. The germs of those distinctions in Style which concern
the harmonic (and, to a certain extent, the rhythmic) details of
a musical sentence, are enumerated in par. 14 and 15, and have
been unavoidably involved in all the foregoing Exercises. But
style, in the broader sense of the term, depends upon certain
general characteristics, to which the attention of the student must
now be directed ; namely :

(a) Upon the choice of Time ; i.e., the use of the Duple or
Triple rhythmic group (as respectively embodied in the I or * Time
of the March, etc., and in the J or | Time of the Waltz, Barcarolle,
etc.). The characteristic effect of each of these two fundamental
classes is so generally appreciated, that it suffices here to observe
that duple measure is the more vigorous, and triple measure the
more graceful species. There are no other species, unless the
peculiarity of | (or J) Time, and of \ (or ]) Time, be regarded
as a radical metrical difference.

Par.97d. STYLE. T75

(b) Upon the choice of Tempo; i. e. rate of speed, from
Adagio^ through Largo^ Larghetto^ Andante* Allegretto^ Allegro
and Presto^ up to Prestissimo,

The deep significance of this distinction will appear upon comparing
Mendelssohn, ** Song without Words '^ No. 22, with No. lo^ and these again
with No. 36; or the 4 Movements of Beethoven's Pfte. Son. op. 2, No. 2, one
with another.

In indicating the tempo^ it should be borne in mind that the given rate —
for instance, Allegro — refers to the heat in the adopted Time-signature : in
I Time it must mean that the 8th-notes pass at an allegro-ra.te of speed ; in J
the same rate would apply to the quarter-notes. The uncertainty which
prevails, because of a general misunderstanding or violation of this rule, can
be dispelled by employing the metronome-marks, — e. g. ^ or Jor J = 6o, etc.

(c) Upon the choice of principal Mode, i.e., major or minor,
the characteristic distinctions of which are doubtless sufficiently
palpable to every musical sense.

(d) Upon certain Rhythmic peculiarities, which, while
available at any time and in any consistent connection, have
become chiefly characteristic of certain conventional varieties of
composition ; for example, the respective typical rhythms of the
Polonaise, the Bolero, the March, Minuet, Mazurka, etc.

It would, however, be a deplorable limitation of the student's resources,
were he to confine his experiments in ** rhythmic style" to those popular
forms of which a certain (permanent) rhythmic figure is the ruling element.
Unity, on the one hand, will dictate close adherence to the adopted rhythmic
figure throughout the section, and even on to the end of the Song-form, — as
in Mendelssohn, ** Songs without Words'* Nos. i, 8, 14, 21, 23, 36, 43; but
Contrast, with its equally just and persistent claims, must not be disregarded,
in planning and executing the rhythmic design. Of the ** Songs without
Words '' cited above, Nos. 8, 21, and 23 bear more or less frequent and striking
traces of the writer's intention of occasionally abandoning the predominant
rhythmic motive.

See Beethoven, Symphony No. 7; observe how, in each one of the
4 Movements, the prevailing rhythmic figures fix the general character or
style. Also Schubert, Fantasie, op. 78, 3rd Movement (J/r«Mc//<7); Momens
musicals^ op. 94, No. 2; No. 4, **Trio" (5-flat signature); No. 5; and ponder
upon the resources of interest and variety suggested by the following
manipulations (in different Divisions^ however, be it understood) of a primary
rhythmic figure:

*^* Andantino** should be used with discrimination. For the definition
of this frequently misinterpreted term, see Baker, " Dictionary of Musical
Terms."

176

STYLE.

Par. 97d.

z. Adagio*

Allegro.

93.

1^

4— g-#-

f~r"] p9^^^

Origiual form.
Mennetto tetn^o.

later :

Allegretto.

later:

Allegretto.

Beethoven, op. 34.

later:

^-^f^

-h "-mJ-

y

later :

2. Andante con moto.

g^ ffe^p^^

Orisciiial form.

.<4 Uegro.

^=i

5=^

^i

later :

Moderato,

Maestoso.

*=^n^3

t=j

#— *-

t=t

15>-T-

E^S

later :

Lento.

later :

I

g=»-"- ^ ? :J^ §^==^

tt

i^

^

^

later :

Mazurka.

^

-H-^L^^

^ llegro.

TSCHAIKOWSKY, op. 50.

ltd:

^^-<a-

i

later :

later :

The systematic demonstration of this fertile rhythmic process must be
deferred until the pupil has undertaken the study of ** Applied Counterpoint;"
but he is already entitled to such uses of it as his stimulated imagination may
achieve. (It is somewhat akin to the processes described in par. 29c and 32.)

There is one general rule, in reference to rhythmic style, which
must be respected, namely :

The less striking (i. e.,less irregular) a rhythmic figure is, the
longer may it be adhered to without incurring monotony ; and,
inversely, the more irregular a rhythmic figure is^ the sooner
should it be exchanged for a simpler Jigure.

Par. Oie. STYLE. 1 77

(e) The well-nigh infinite combinations of the above distinc-
tions, including those given in par. 95 and 96, afford that complex
of resources out of which the scarcely definable (for convenience
called emotional) elements of musical thought are evolved.
These resources cannot, and should not, be theoretically classified ;
it is the duty of the individual student to investigate them for
himself and develop his taste and individual style according to
what he may discover*

But against one very popular, and unquestionably mischievous,
fallacy he must be earnestly warned, namely : the error of supposing
that Music can definitely indicate any object^ tell a story ^ express
any idea clearly^ portray any incident^ or transmit emotional con-
ditions ivith even approximate distinctness. This is not the
MISSION OF THE TONE-ART. And insisting upon its use in the
service of other than intrinsically musical ideas, is a degradation
of its ethereal and lofty calling.

At the same time, the student must quite as fully realize what Music
can do: Music can arouse and can reflect emotional states, sometimes with
greater power, and always with more searching keenness, than other arts. But
it is not possible to foretell ivhich emotional state will be aroused, save in an
extremely general way; and the subtility of musical impressions is the very
attribute which distinguishes the emotional magnetism of Music from all
other emotional stimuli, and robs these impressions of all definiteness.

Further, Music can also roughly delineate and imitate (by its melodic
** lines,'' its rhythmic ** movements," etc.) certain primary motions of the
material world; but with only just enough clearness to suggest^ — and only
suggest, — any one of a thousand similar material motions.*

This imposes thje undeniably binding duty upon the composer, of so
choosing these accidental elements of suggestion as to avoid incongruity.
For example, in a ** Slumber ''-song constant or frequent fortissimi would
be inappropriate, — except as an exposition of '* Humor in Music." For an
exquisite example of appropriate musical suggestiveness, see Schumann,
op. 15, No. 12 (** Child falling asleep "); especially the last 6 measures, in which
the extraordinary widening-out of the chord-forms suggests the distorted
fancies of half-slumber, while the tired little soul floats out into oblivion upon
the unresolved J chord of the Sub-dominant. He who can play or hear this
passage without deep emotion, has not learned the power of the gentle Art
of Tone !

The pupil need make no definite effort to apply these general
truths. He should ** bear them in mind" and ponder over them,

*See the Revue philosophique for Febr., 1893, pages 124-144 (translated
in the ** Music Review/' July and August, 1894).

178 CODA AND CODETTA. Par. 98a.

and leave the issue to the absorbing and assimilating processes
peculiar to his own mental organism. In the meantime he is to
follow all the given rules to the very letter.

Coda and Codetta.

08. The difference between these two terms is solely one
of extent, — '* Coda " being the original designation for an indepen-
dent ending, and ** Codetta" signifying '*a small Coda." The
distinction may be formulated with greater accuracy by calling
every single addition at the end (even when repeated and
extended) a codetta, and defining the coda as a series of

CODETTAS.

The latter (a series of Codettas) is extremely unlikely to appear anywhere
in the course of a composition ; hence it is saf« to say that the term Coda will
be applied only to the ending of the whole, whereas a Codetta may be added to
any sufficiently definite member or Part^ even in the course of the design.

(a) The object and character of the Codetta are defined in
par. 51, which carefully review, with all the illustrations; and see
also par. 27 and par. 32. — ^The primary purpose, which corresponds
to the ** Extension of a Phrase-cadence," remains the same through
all the grades of dimension in which it is achieved ; and the
dimension, or length, of the Codetta will increase in a more or
less regular ratio to the size of the body of the design to w^hich
it is affixed.

These proportions may be approximately graded, as follows, —
it being remembered that a complete perf, cadence usually precedes
the Coda or Codetta^ to mark the limit of the body of the design.

(i) The shortest cadence-extension is the reiteration (perhaps
3, 3, or 4 times) of the single final Tonic chord,

(2) The next larger extension consists in a repetition (perhaps
2 or 3 times) of the two cadence-chords (V— I).

(3) The next larger formation is somewhat apt to be a brief
Codetta (i. e., not an extension of the given cadence, but an
** appendix," independent of the latter), about two measures in
length; generally repeated, and possibly extended as at (i) or (2).

(4) The next larger independent ending would be a Codetta
oi four measures^ probably repeated and extended.

Par. 08b. CODA AND CODETTA. 1 79

(5) The next larger formation might be such a 4-measure
Codetta (repeated), followed by another of 2 measures (repeated
and extended). This series of Codettas would be termed a Coda.

(6) A still broader Coda would consist of an 8-measure
Codetta (possibly repeated), followed by one of 4 measures (prob-
ably repeated) , and another of 2 measures (almost certainly repeated
and extended). No larger form of Coda would be appropriate in
the present grade of forms.

Nos. 2 and i of this schedule will suffice, as a rule, for the
*' Phrase"; Nos. 3, 2, and i for a large Phrase, or ''Period";
Nos. 4, 2 and i for a large Period, Double period, or ''Part" ; Nos,
5, 2 and I, or No. 6, for an entire 2-Part or 3-Part " Song-form."
It will be observed that the proportionate length is obtained by
adding a higher (larger) grade, and not by singling out any
special grade ; in other words, the smaller grades are all expected
to follow, in order to preserve the tapering form. But it must
be distinctly understood that this entire schedule is only an approxi-
mate theoretical tahle^ and not necessarily practically bindings save
as it illustrates an important and generally valid principle.

An exact illustration of this tapering form is found in Chopin, Mazurka
35 (op. 56, No. 3), last 32 measures (Coda of 5 sections: an 8-meas. section,
with repetition; then a 4-meas. Phr. with repetition; then a 2-meas. Codetta,
with repetition; then one measure, repeated, and a final reiteration of the last
chord).

Chopin, Nocturne No. 8 (op. 27, No. 2), last isJjj measures (4-mea8.
Codetta, repeated ; 2-meas. Codetta, repeated ; Tonic extension).

Mendelssohn, ** Song without Words " No. 30, last 19 measures (4-meas.
Codetta, repeated; 2 measures, repeated; one measure, 4 times; final chord
through 3 measures) .

Beethoven, Bagatelle op. 33, No. 2, last 28 measures (similar); Bagatelle
op. 33, No. I, last I3J<^ measures (4 measures, repeated; 2 measures, repeated
twice; one measure); Bagatelle op. 33, No. 6; see also op. 33, Nos. 3, 4 and 7.
. Mendelssohn, ** Songs without Words," No. 11 (last 8^^ measures); No.
12 (last 12 measures); No. 14 (last 11 meas.); No. 19 (last 15)^ meas.); No. 34
(last 17 meas.); No. 42 (last 15 meas.).

In the "Songs without Words," Nos. i, 13, 25, etc., the entire **Song"
has no more than a brief Codetta. In No. 24 the Coda is very long (44)^
meas.). In No. 6 there is first a Codetta, and then a Postlude corresponding
to the Prelude. In Nos. 7, 8, 10, etc., the formation of the Coda is irregular,
1. e., not agreeing with the above schedule.

(b) 'It has already been stated that a Codetta, being somewhat
independent of the body of the design, may also be independent in

l8o CODA AND CODETTA. Par. 98c.

its thematic structure. This is a little more likely to be the case
with larger Codettas, and Codas, whereas the smaller appendages
are less likely to contain new motives. The following examples
will sufficiently illustrate the derivation and thematic character of
the Coda :

Mendelssohn, ** Songs without Words " Nos. 3, 6, 7, 14, 21, 25, 26, 31, 37 ;
here the Coda or Codetta is in each^case derived from the first Phrase of Part
I ; often by ** Melody-expansion " (par. 32). In No. 15 it is developed out of
the end of Part III.

In Nos. 29, 32, 44, 46 it is a reflex of Part Second. This is more logical
than it is usual, in composition; for the Coda will follow Part III most
naturally in the same way that Part II followed its First Part ; and Part II,
if independent of Part I, may need just such corroboration (see par. 103).
It will be observed that in this case the melodic member used will appear in
two different tonalities ; at first in the key or harmony of the Second Part, and
in the Coda, — of course, — transposed to the principal key.

In Nos. 22 and 43, and in op. 72, No. 4 (Mendelssohn), the Coda is
a contracted recapitulation of Parts II and III. In Nos. 11, 12, 19, 36, 48,
it contains more or less new melodic material.

(c) The most appropriate place for a Codetta is at the end
of the entire structure, where, according to the dimension of the
body, it may, if necessary, expand into a Coda.

But the possibility of thus rounding off an earlier section of
the form has been touched upon, and examples of a Codetta to the
Second Part have already been seen (par. 89^, last clause).

Far more natural and customary than this, however, is the
addition of a brief Codetta to the First Part^ usually w^ith repeti-
tion and extfension. This presupposes that the design of the entire
Song- form is somewhat broader than usual.

See Mendelssohn, "Song without Words" No. 39, last 2 measures of
Part I.

Haydn, Symphony No. 3 (Peters ed.), Menuetto^ Part I; Symph. No. 4,
MenuettOy Part I ; Symph. No. 8, MenuettOy Part I; No. 9, Meuuetto^ Part I.

EXERCISE 26.

A, An example in major, (ordinary Complete 3-Part Song-form as in

former chapter,) with such application of the irregular Cadence-conditions

(par. 93) as may be appropriate ; with greater modulatory freedom (par. 94) ;

and with some dynamic design^ — the Song to contain at least one // and one

ff^ disposition optional (par. 95) . A Coda is to be added (par. 98) .

Also a similar example in minor, with Coda.

Par. 09a.

THE INCOMPLETE 3-PART SONG-FORM.

181

B, A similar example in Adagio tempo (par. 97^), with some regard to
the conditions explained in par. 96 ; and with a Coda,

An example in Presto tempo, with a Codetta to Part I, and a Coda to the
whole.

C Optional, (but simple 3-Part form). See par. 137,

CHAPTER XIV.
The Incomplete 3-Part Song-form.

99. In the Incomplete (or abbreviated) form, the Third Part
IS considerably shorter than Part /, in consequence of reproduc-
ing only a portion, instead of the whole, of the latter.

It is evident that this reproduced portion must, however,
include the very jirst memher of Part I, at least, in order to fulfil
the essential condition of the tripartite design, i. e., the return to
the beginning (^zx . 8ia; par. 84^, 4th clause). A single measure,
if of somewhat striking character, may be sufficient to establish
this distinction ; but, as a rule, no less than an entire Phrase of
Part I recurs, as Part III.

(a) If the First Part is a Period of parallel construction,
Part III may appear to be the Consequent Phrase ; or it may be
combined out of the essential members of both Phrases. But
if the construction of Part I is contrasting, or if it consists of
a Group of dissimilar Phrases, then the^r^^ of these Phrases must
constitute the body of Part III.

The difference between the "Incomplete** and the "Incipient" 3-Part
Song-form is discoverable in the extent (or form) of the Second Part. In
the Incipient form, both Part III and Part II are so brief as to constitute
together, apparently, no more than the equivalent of Part I ; in the Incomplete
form, on the contrary, Part II is a full-fledged '* Part,'* about the identity
of which there can be no question, not even of quantity, while Part III is
a decidedly contracted recurrence of the First Part. Thus :

Phr.

Phr.

Phr.

II VlIl(I)

Phr. Phr.

II V III (I)

Per. or Group
_

optional

n

= the 3-Patt Period.

= the Incipient 3-Part Song-form.
Phr.

V III (I)

Period

"V III (I;\r"

= the Incomplete 3-Part Song-form.

= the ordinary Complete form.

l82

THE INCOMPLETE 3-PART SONG-FORM,

Par. 99a.

For illustration :

Assai tiUegro.

84.

a¥f I ^ ^rir

Partn.

ff

N=^ ^ Jjjl^ l

^

^

y ^>

i

I

i

/^v

Part in.

tfe

#=?:

^^

" ^ - ti-i-i

±

i

•-'//Tn P

^^

■/^-rg: ^

^i

«

*

2k

i^

Beethoven.

^

I

r-y^

^

Par. Wc. THE INCOMPLETB 3-PART SONG- FORM. 183

*i) The First Part is a full parallel Period, of 8 measures; Part II is also
S measures in length; Part III contains only -six measures, in Phrase-form.
At first glance, 3 or 4 of these last measures may seem to be derived from
the Consequent of Part I ; but, while they correspond to the latter in register
only, their thematic agreement with the first Phrase is complete. The 5th
measure is derived from the last measure of Part I, and a 6th measure is added
for the sake of symmetry.

See also : Mendelssohn, op. 82, Theme ; op. 83, Theme ; op. 72, No. 4 ; —
in each case Part III is just one-half the contents of the First Part; apparently
the 2nd half, but always including the beginning.

Mendelssohn, ** Songs without Words '* Nos. 12 ; 20; 15 ; 43. In No. 3,
the First Part is a Group of six Phrases; Part III contains the ist, 2nd and
6th of these Phrases, besides 3 other (intermediate) measures.

Chopin, Mazurka No. 11; No. 32, first 44 measures; No. 51, first 40
measures (Part II an exact reproduction of Part I, but in a different key).
Chopin, Nocturne No. 3 (op. 9, No. 3), ^^ Agitato^^^ — Part III derived from
second half of First Part, but including the beginning.

Mozart, Pfte. Son. No. i (Cotta ed.), Andante, (Part III corresponds
exactly to 2nd half of Part I, but the construction of the latter is parallel.

Haydn, Symphony No. 5 (Peters ed.). Finale, Principal Theme (model
Second Part, sectional form; Part III like 2nd half of Part I, but including
beginning).

Schubert, Pfte. Son. No. 6 (op. 147), **Trio" of 3rd Movement.

Beethoven, Pfte. Son. op. 2, No. 3, Scherzo (Part III about one-half the
contents of First Part, and followed by two Codettas of 8 and 9 measures).

Schumann, op. 82 (** Waldscenen"), No. 5.

Grieo, Lyric Pieces, op. 43, No. 2 (II and III repeated).

Brahms, op. 116, No. 3, •* un poco meno AllegroJ^ — Op. 118, No. 2, up to
first double-bar (Part I a reg. Period, repeated ; the first member of Part III is
in the '* contrary motion " of Part I ; Part III a Phrase, followed by Codetta,
rep. and extended). Op. 118, No. 4, up to double-bar (both Part I and Part III
close with a semicadence). — Op. 119, No. 2, second tempo (4-sharp signature).

(b) In some cases the Third Part is only a slightly contracted
version of Part I, — ^as, for instance, in Ex. 91 (only one measure
omitted). Such examples, where Part III does not omit any essen-
tial portion of Part I, should be classed among the Complete forms.

This may seem to place Ex. 94 in a somewhat doubtful position, though
its Third Part is an indisputable contraction in /br/», from Period to Phrase.

See also Chopin, Nocturne No. 19 (op. 72, No. i) ; Part I, 21 measures.
Part III, 16 measures.

Mendelssohn, Praeludium op. 35, No. i (Interlude betw. Parts I and II ;
elaborate Coda).

(c) On the other hand, as already intimated, a very few of the
first beats of Part I, if fairly striking in character, will suffice to
establish the tripartite design. But, unless this meagre indication

184

THB INCOMPLETE 3-PART SONG-FORM.

Par. 99c.

of a ** Return to the beginning " be followed by enough consistent
material to fill the just measure of a full-fledged **Part" (in the
manner to be seen in the 2nd and 3rd stages of the fully developed
form, par. 102^ and c), it will not, of itself alone, entirely over-
come the impression of a 7w<?-Part design. For illustration :

Aftdante,

^

I

I:

g

m

etc.

i

^^^^

^

Part n.

Y

fjj^UM' riti^

(iTTTn

etc. See originaL

iiiia

I:

I I I

a

V

^£^^^^t^

#=

W^^

P

Crete.

^il:

Part III. *I)

. .. nTn m Jl

rr tr

rATm UtfCa

THK INCOMPLETE 3-PART SONG-FORM.

X85

Mbmdblssohn, op. 7, No. 6.

*i) There maj be some doubt as to whether this last 4-measure sentence
suffices to represent a definite Third Part, but the doubt is scarcely reasonable.
It contains, to be sure, less than one-half of its First Part ; but, firstly, it is
a complete 4-meas. Phrase; and, secondly, the peculiarly effective dynamic
conditions with which it is ushered in {cresc, molto ^^^ //), emphatically
mark one of the vital "angles" of the structural design. Almost precisely
the same conditions prevail in the same opus (7), No. i.

In Mendelssohn, op. 14, Andante (Introduction to Rondo), the recur-
rence of the first member, 5 measures from the end, is barely more than an
intimation^ and therefore the form is probably only 2-Part.

In Chopin, Prelude, op. 28, No. 21, the intimation of the Recurrence
is also vague, but the form is surely 3-Part, on account of the individuality of
the Second Part. In the following Prelude (No. 22), a Third Part can hardly
be proven.

In Beethoven, Symphony No. 8, ** Trio " of 3rd Movement, there is
a fairly convincing intimation of a Third Part, in meas. 9 of Part II.

' In Haydn, F-minor Pfte. Variations, the intimation of a Third Part
is very vague in the Principal Theme, but quite distinct in the F-major
Division which follows. Consequently, the former is probably 2-Part, the
latter Incipient 3-Part form.

In Bach, Well-tempered Clavichord, Vol. II, Prelude 12, a Third Part is
** intimated" ; the same in Bach, English Suite No. i, Sarabande,

In Schumann, yugend- Album ^ op. 68, No. 7 (F major), the intimation
is so vague that the form is obviously only 2-Part; the same is true of
Mendelssohn, Scherzo a capriccio (FJ(), Principal Theme; Beethoven,
7 Variations in F (P. v. Winter), Theme ; Haydn, Symphony No. 6 (Peters
ed.) Andante (32 measures; 2-Part form, each Part repeated).

EXERCISE 27.

Two examples, one in major and one in minor, of the Incomplete 3-Part
Song-form. — Choose a different kind of Time, grade of Tempo, dynamic
design and rhythmic style for each. Review the directions given in Exercise
9, last clause. — Add a Codetta or Coda, and an Introduction, if appropriate.

1 86

THE AUGMENTED TWO-PART SONG-FORM.

Par. 100.

The Augmented Two-Part Song-form.

1 00. The occasional inclination of a Second Part to revert,
in its later course, more or less emphatically to the first member of
Part I, as seen in the preceding paragraph, is scarcely less likely
to involve some other ^ later ^fragment of the First Part,

If the fragment thus borrowed from Part I is any other than
the first member, every suspicion of the genuine tripartite form
vanishes, because it is no longer a return to the beginning : the
form is unquestionably Two-Part.

If, again, the borrowed fragment is merely a corroboration
of the ending of Part I, as ending of Part II, without causing
any palpable check in the current of the latter, there is quite as
evidently no essential change or enlargement of the Two-Part"
design (see Ex. 73).

If, however, finally, the Second Part has been sufficiently
individualized, arfd, in consequence, a disposition to '' return "
asserts itself, but, in thus evidently returning, choice chances to
fall upon some later (though equally characteristic) member of the
First Part, it is certain that the form has become richer than
the ordinary Two-Part design, though without reaching the lowest
limits of the Three-V^iXX. form. (Review par. 65, and par. 77,
first clause.) For this intermediate design the term "Augmented
2-Part form " seems most appropriate.

The form is strictly only bipartite, because the vital distinction of the
tripartite form is wanting, excepting as its presence is to a certain degree
** understood '' through the act of substitution which takes place, whereby
an inferior agent is accepted as the representative of the expected one. If
** distinct evidence of a return to the beginning''^ is to be upheld as an
inviolable condition of the genuine Three-Part Song-form, then the follow-
ing Example (96) is only a bipartite design ; and such other, still larger
specimens as maybe found, consisting unmistakably of three distinct Parts ^
but lacking this evidence of a return to the beginning, must be classified
among the *' Group-forms*' (par. 114). For illustration:

A dagio.

96. <

Par. 100.

THE AUGMENTED TWO-PART SONG-FORM.

187

k U f

1/

*i) Old-German love-song (harmonized by L. Stark).

*2) Part II is a full Period, up to this point, equivalent in length and
form to the First Part. The Phrase which follows is not derived from the
beginning but from the ending of Part I, and therefore it is not a genuine
Third Part. Under other circumstances, such an example as this, especially
if broader in dimensions, might be called a Group of (3) Parts ; but, consider-
ing the close affinity between this final Phrase and all the rest of the Song, it
appears most accurate to speak of it as an Augmentation of what is actually
no more, in substance, than a Two-Part form. The confusing resemblance
between this example and Ex. 73 cannot be denied ; but there ^ the Second Part
reverts to the ending of Part I without checking its course, or augmenting
its formal design ; while herey the Second Part might seem complete without
the addition ; therefore the latter is really an augmentation of the design.

An excellent illustration is to be found in Mendelssohn, ** Song without
Words " No. 18 ; in measures 30-32 the Second Part evinces all the symptoms
of coming to an end, and in meas. 32 a motive sets in (with strong dynamic
emphasis and every other indication of a "new Part") which probably any
one, at a first hearing, would accept as a genuine Third Part ; — on examination
it proves to be the second member of Part I, and not the initial one at all.
The formal proportions are almost exactly the same as in Ex. 96, so the same
reasons prevail for calling it an Augmented Two-Part form.

See also Brahms, Rhapsody ^ op. 79, No. i, ** Trio " (signature of 5 sharps).
In meas. 8-9 of the Second Part there are strong indications of a regular
Dominant-ending ; but the Phrase which follows is not sufficiently convincing
of a return to the beginning.

Brahms, op. 116, No. 6 (Intermezzo) first 24 measures; Part II contains
a reproduction of all the First Part, excepting its first Member (2 measures) ;
one is tempted, therefore, to call this ** 3-Part Song-form." Brahms, Pfte.
Ballad, op. 10, No. 2, first 23 measures (excellent illustration).

See also Chopin, Nocturne No. 3 (op. 9, No. 3), Principal Song (Part I
20 measures and repeated ; Part II 16 measures, followed by an Augmentation,
consisting in the last Phrases — 8 measures — of Part I).

In Chopin, Mazurka No. 28, the design must be called Incomplete 3-Part
Song-form, because the first member of Part I — omitted in Part III — is much
like an unessential Introduction ; and because every other essential condition
points to three distinct Parts^ the third of which is an almost exact recur-
rence of Part I, contracted, and without the ** Introduction."

l88 THE FULLY DEVELOPED THREE-PART SONG-FORM. Par. 102b.

EXERCISE 28.

An example of the Augmented Two-Part form, all details optional.

CHAPTER XV.

THE FULLY DEVELOPED THREE-PART SONG-FORM.

101. The complete development of the resources of the 3-Part
Song-form depends, finally, upon the treatment of the Third
Part, which, after having fulfilled the necessary condition of the
tripartite design by reproducing the Jlrst member of Part /, may
thereafter be elaborated into a more or less individual Part, inde-
pendent enough (in its later conduct) of its original First Part, to
constitute a coordinately interesting and essential section of the
form. In this line of progressive development, four stages may be
distinguished.

1 02d.. Stage i is represented by Ex. 91 and its context
(par. 91) ; here Part III is either a literal recurrence of the First
Part, or contains no other than such unessential modifications, and
3r/^ extensions, as add to its attractiveness without affecting the
form, or any other essential feature of the Part. (Mendelssohn,
*' Songs without Words," Nos. 45, 22, 32, 35.)

(b) In Stage 2, the Third Part is an extended version of its
First Part, or, more commonly, of a portion of the same. As
a rule, it reproduces only a certain fraction or section of Part I
literally; generally no more than the first Phrase, or such a portion
of the "beginning" as suffices to establish the tripartite design ;
and then, after doing this, fills out the remainder of its allotted
length (usually at least equal to, if not beyond, the length of
Part I) by an extension or expansion of this member^ along lines
approximately parallel with its First Part. For illustration :

Andante,

97.

Par. lOeb. THE FULLY DEVELOPED THREE-PART SONG-FORM.

189

Part II.

I

^

r-*-

^^^^

gfe

Part in.

f=>i

• •

^M

t^

8 measares;
see Origiual *1)

m

• •

^

*i) Haydn, Symphony No. i (Peters ed.), Andante, The principal
melody of Part I is transferred, in Part III, to the lower part, and is retained
almost literally for 4 measures.

*2) From here on. Part III abandons the line of the First Part, and carries
out an interesting extension of the melodic figure of its 4th measure, followed
by a somewhat independent Cadence-member. The Third Part exceeds Part I
in length by 2 measures.

See also: Haydn, Pfte. Son. No. 9(Cotta ed.) Presto^ Prin. Theme.

190

THE FULLY DEVELOPED THREE-PART SONG-FORM. Par. 102c.

Beethoven, Pfte. Son. op. 7, 3rd Movement, up to the *' Minore " (Part III
very elaborate, — perhaps passing beyond this 2nd Stage; a Codetta is added);
the same Sonata, Largo^ Prin. Theme (similar).

Mozart, Pfte. Son. No. 9 (Cotta ed.) **Trio'' of Menuetto (peculiar
example; Part III contains no more than the first ij^ measures of Part I,
exact ; the rest of the way — 15 measures — not a single member agrees literally
with the original Part, but the general parallelism is so perfect that the actual
difference would not be suspected).

Chopin, Mazurka 18; Nocturne 18 (op. 62, No. 2), 32 measures (also
Incomplete Form, i. e., Part III decidedly shorter than I); Nocturne 13 (op.
48, No. i) Prin. Theme (somewhat beyond the 2nd Stage).

Mendelssohn, ** Songs without Words,*' No. 42; 46; 31; (compare, in
every instance, Part III very minutely with Part I).

Grieg, Lyric Pieces, op. 43, No. 5.

Brahms, op. 117, No. 2; op. 118, No. i.

(c) Stage 3 is characterized by the presence of more or less
radically new melodic material in Part III. It emerges naturally
out of the 2nd Stage, and is sometimes scarcely distinguishable
from the latter, because it is not always possible to determine just
when the * * extension ' ' of the initial portion of Part I digresses
far enough from its original object to become *'new" in effect.
But usually the evidence of really new (though, of course, strictly
kindred) members in the Third Part is unmistakable, as in the
following :

Allegretto. ^^ ^^ Y

I

^f JTw^ J

rt^mm

PartIL

i

-8 measnres;
Isee Oiiginair

Part m. (from Part I.)

I

^m

Y *1) (Extensionof prin. member.).

\ — H

^#

• •

X

erete.

y (New member).

y (Reproduction of new member). *2)

Mrndblssohn, op. 73, No. 3.

eto. Codetta.

Par.lCWd. THE FULLY DEVELOPED THREE-PART SONG-FORM. I9I

*i) Up to this point, Part III is a literal reproduction of the first Phrase
of the First Part; then the line of Part I is abandoned, and in its place stands,
first, a 4-meas. extension of the first figure, and then an entirely new, but
sufficiently homogeneous, member, which is confirmed by reproduction.

*2) In connection with this reproduction of the new member, see par. 103.

This 3rd Stage represents the most perfect, richly developed,
and beautiful variety of the genuine 3-Part Song-form ; and it is,
therefore, quite commonly adopted, especially by modern writers.
Examples of it are very numerous, but its resources are not, and
cannot be, exhausted. (Stage 4, as will be seen, contains one
further trait of structural coherency, but is, on the whole, a trifle
less genuine than the 3rd Stage.)

See also: Mendelssohn, ** Songs without Words," Nos. 37, 40 (between
the 2nd and 3rd Stages); Nos. 25, 26, 36, 47 (all 3rd Stage). In No. 2, the Third
Part has a somewhat suspicious break in its i6th measure, upon an imperfect
Tonic cadence, but the following 16 measures are so intimately related to the
preceding members, that they can hardly be called a Coda (comp. par. 93^).

Schubert, Pfte. Son. No. 6 (op. 147), Andante, Prin. Theme.

Chopin, Mazurka 32 (op. 50, No. 2) last 100 measures (comp. with Principal
Song, at the beginning), — Part I 16 measures. Part II 16 measures, Part III
33 measures, 17 of which are '* new.'*

Beethoven, Pfte. Son. op. 10, No. 2, Allegretto, first Song (codetta added);
Pfte. Son. op. 10, No. 3, Menuetto, first Song; op. 14, No. i, Allegretto, first
Song; Symphony No. i, ** Trio " of 3rd Movement; Symphony No. 3, ** Trio "
of 3rd Movement.

Brahms, op. 118, No. 3, Subordinate Theme (5-sharp signature). This
example might be called a Quadruple Period, but the 4th Phrase is so distinct
in style as to tempt the assumption of a ** Second Part," brief and Interlude-
like though it is. Op. 119, No. 3 (Part I 24 meas.. Part II 16 meas.. Part III
modified at beginning, and extended). Capriccio, op. 76, No. 8.

Schubert, Motneus musicals, op. 94, No. 6, first Song.

Schumann, op. 82 (** Waldscenen") No. i; Part III very elaborate and
followed by a Codetta, after an imperfect Cad. on a J chord.

(d) In Stage 4, the Third Part contains, besides its own
rightful material, some characteristic member of the Second Part,

The portion thus borrowed from the (foregoing) Second' Part
may be of any extent, from a single melodic member up to a
Phrase, or an entire Period; but it must be something peculiar
to Part 11^ and not a passage which found its way into that Part
from Part I, — ^as is more likely to be the case ; hence, this device
is more liable to be adopted when Part II is partly **new" in
construction (par. 87^). In no case will Part III include the whole
of Part II, but only some characteristic portion.

192

THE FULLY DEVELOPED THREE-PART SONG-FORM. Par. 102d.

The borrowed member will almost invariably be subjected to
a transposition ; for, while likely to appear at first in some key
conforming to the modulatory design of the Second Part, it will, in
Part III, be transferred to the principal key.

Finally, the member borrowed from Part II, must, in order
to affect the design, recur during t^le course oi the Third Part, i. e.,
before its final cadence. Otherwise it will only become a part
of the Codetta or Coda (as seen in par. 983, last two clauses, which
review). For illustration :

A ndante.

^

^

m-

i

1.

¥a.

PartIL

3=t=:|:

^v^ ^ 1 >

:?2— ;5=

-— *

■XT

H

m

urn.

tranquiUo.

i kh-^:^ L ^ \ ?:;^^^n[j^ ^^

^S

Conseq. Phr.

fH d r^

Crete,

/ *"** .,/ ' ***" dim, '

3

erete.

f sf dim. pocoraU,

Part m. (from Part I).

i feZJ ? J . i^

Z

■Jt=t

r J ^ L4

d^ ^ > ^ '^

u i r r rr

V

/

♦1)

Par. 103. CORROBORATION. I93

( from Part 11) Mendelssohn. S. w. W., No. 7.

rep. &Goda.

*i) Up to this point, Part III pursues the line of the first Phrase of Part I
closely; what follows is neither a continuation of Part I with unessential
changes (as in Stage i), nor a mere extension of the foregoing member (as in
Stage 2), nor is it new (as in Stage 3), but is borrowed from the foregoing
Second Party to the Consequent Phrase of which italmost exactly corresponds,
save in key. This is an unusually striking example; so characteristic, in
fact, that the Third Part might be called a contracted recurrence, or '* Re-
capitulation," of both Parts I and II. Compare it carefully with the text
(par. io2tf), all the conditions of which it fulfils.

See also : Mendelssohn, '* Songs without Words,'* No. 41 (similar to the
above; Part III further enriched by an additional **new'* member); No. 38
(meas. 10-19 of Part III are derived from meas. 6-8 of Part II) ; No. 30 (meas.
15-22 of Part III borrowed from Part II); No. 13 (meas. 13-15 of Part III
taken from II) ; No. i (meas. 5-6 of III borrowed from II). In Nos. 10, 33, 28
the coincidence between portions of Parts III and II is more general than
detailed. — ^Mendelssohn, Praeludium op. 35, No. 5. In Praeludium op. 35,
No. 4, there is a suspicion of the borrowed portion constituting the beginning
of the Coda. See also op. 7, No. 2 (very broad; large Second Part) ; Andante
cantabile in B major.

Chopin, fitude No. 2 (" Moscheles-F6tis " Method).

Brahms, op. 116, No. 4. — Schumann, Jugend-Alhuniy op. 68, No. 2. —
Beethoven, Symphony No. 2, Scherzo^ Principal Song.

Corroboration.

1 03- According to the law of Corroboration, — which be-
comes more imperative in proportion to the increasing dimension
of the musical design,— every member which assumes a sufficiently
striking or impressive individuality to attract more than passing
attention, should recur more or less exactly, sooner or later, (it
does not much matter when or where,) for the sake of thematic
confirmation and structural balance. Or, negatively expressed, it
is necessary to avoid the isolation of any conspicuous passage or
tnemher; the law of Unity demands its corroboration somewhere
or other within reasonable limits.

This rule is involved by the law of Contrast (par. 96, which review), the
action of which must be counterbalanced by that of Corroboration. The
greater the Contrast, — ^be it in quantity or quality, — the greater the need of
Corroboration; hence, these two equally important conditions of rational
Form are mutually dependent.

194 CORROBORATION. Par. lOS.

This very strongly emphasizes the principle of Repetition, and
of Reproduction, as fundamental structural law; see par. 17, 23,
Ex. 46, par. 41, 42, 43, 54, etc.

It demonstrates the logical justice of the tripartite designs ;
see par. 8ia, 91.

It accounts for the occasional similarity between the ending
of Parts I and II (par. 75).

It is often the direct cause of the Third Part reverting to some
characteristic motive of Part II (par. i02d) ; or of the derivation
of the Coda from Part II, w^hen the latter is somewhat independent
of its First Part.

And, finally, it resolves itself into a more general (though not
rigid) rule for the treatment of any single member which, being
nev)^ or striking^ (in obedience to the demands of Contrast,) rises
abovfe the level of its surroundings and becomes a salient feature
of its Division ; such a member should, in well-balanced Form,
reaffear somewhere, further on, — perhaps in the Coda.

For illustrations of this law, see Ex. 98, Note *2).

Mendelssohn, ** Songs without Words," No. 4 (the independent Prelude,
recurring as Postlude, at the end) ; No. 38 (the independent rhythmic figure
in measures 6 and 7 of Part II, reappearing in Part III).

Beethoven, Rondo in C, op. 51, No. i (Cotta ed., wherein all the Themes
are marked) ; the triplet treatment of the principal motive, in measures 8-9
of the Retransition after the 2nd Subord. Theme, recurs in the following Prin.
Th., measures 4-5; the ascending chromatic run in b-sextolets at the end of
the same Retransition, reappears near the end of the Coda; the a.4cending
arpeggios in measures 1-2, 5-6, of the 2nd Subord. Theme, are corroborated
in measures 6-1 1 of the Coda.

Mozart, Rondo in A minor (Cotta ed. No. 20) ; meas. 5 of the ist Sub.
Th. is developed at the beginning of the ist Retransition ; meas. 16-18 of the
ist Sub. Th. reappear later on in the same Retransition ; the ^-figure in meas.
5-6 of the 2nd Sub. Th. reappears near the end of the Coda; meas. 5-10 of the
2nd Retransition are reflected at the end of the Coda.

Many other proofs might be adduced, of an equally convincing character,
in confirmation of the classic rule: *^ Shun the isolation of any significant or
striking member, ^^

EXERCISE 29.

An example (major) of the 2nd Stage of the Fully Developed 3-Part
Song-form (par. 102^.). Review par. 93 to 97. Add Coda or Codetta, and
Introduction, if appropriate. Par. 130 may be referred to, and utilized if
desirable.

Par. 104. THE LARGE PHRASE-GROUP. I95

EXERCISE 30.

Two examples (minor and major) of the Srd Sta^e (psir, 102c). Review
par. i©3. See par. 137. — Par. 133 may be referred to, and utilized if desirable.

EXERCISE 31.

An example (all conditions optional) of the 4th Stage (par. io2d), — Par.
135 may be referred to, and utilized if desirable.

The Large Phrase-Group.

1 04. Notwithstanding the distrust which the Group-forms
arouse, and the caution enjoined upoh the beginner in their con-
ception and employment (par. 58, final clause), their peculiar
importance among the architectural purposes of musical Form is
undeniable, and their value (if for no other aim than that of
variety, and a relaxation of the rigid lines of the primary designs)
is positive enough, when used in the proper place, and in judicious
proportion.

How easily possible it is to obtain a long" line of Phrases, — for
which the term ** Chain " might be more fitting than *' Group,"
— by adding Phrase after Phrase, without pausing to sever the
connection, and without diverging from the original character to
such an extent as to indicate a new ** Part," may be tested in such
examples as the Pfte. Etudes of Cramer, and other similar com-
positions.

But, the longer the '* Chain" becomes, the stronger will the
inclination assert itself to mold it into a general agreement with
one of the regular designs, and thus to give to the otherwise
almost inevitably incoherent series a more recognizable shape, at
least approaching the design of the Song- forms, or a broadly-
magnified Double or Quadruple period.

See Cramer, fitude Nos. i, 2, 6, 13 (original complete edition, Peters);
— these are simply Phrase-Chains, with little or no indication of any higher
structural purpose or disposition; Nos. 8, 10, 14, 78 (quasi 2-Part Song-form);
No. 5 (quasi Double period) ; Nos. 4, 7 (quasi 3-Part Song-form).

Haydn, Pfte. Son. No. 6 (Cotta ed.) Finale (an introductory Period,
answering roughly for a First Part; then 15 Phrases in Chain-form, followed
by two Codettas).

Chopin, £tude No. i (**Moscheles-F6tis*' method), quasi 2-Part Song-form.

196 . SIMPLE REPETITION OF THE PARTS. Par. lOBa.

Bach, Well-tempered Clavichord, Vol. I, Preludes i, 2, 5, 6, 13, 15, (simple
Chain of Phrases) ; Vol. I, Prelude 10 (probably 2-Part Song, with long and
distinct Coda) ; VoL I, Prelude 20 (quasi 3-Part) ; 21 (quasi 2-Part) ; Vol. II,
Prelude 6 (quasi 2-Part).

Grieg, Lyric Pieces op. 38, No. 5 (quasi Double period; Coda).

Brahms, op. 116, No. 2 (Intermezzo) fWH troppo presto (6 Phrases, quasi
Double period, extended). Brahms, op. 116, No. 4, Part I (7 Phrases, quasi
Double period). — Op. 119, No. 2, first tempo (quasi 3-Part Song- form; a brief
thematic Phrase reappears several times, in various keys and styles).

CHAPTER XVL

THE EVOLUTION OF THE FIVE-PART SONG-FORM.

I. Simple Repetition of the Parts,

, 105. The growth of the ordinary tripartite Song- form into
a correspondingly legitimate form of Five Parts is initiated by
the simple condition of Repetition, applied to the Divisions
of the 3-Part form.

(a) The general rule for these repetitions is as follows :

Part I may be repeated alone ;

Part II should not be repeated alone, but only in company
with Part III.

In other words : the ist Division (consisting of Part I) may be
repeated, and the 2nd Division (consisting of Parts II and III
together) likewise ; thus :

II HI

(See par. 83a.)

Either one alone, or both, of these Division-repetitions may
occur ; and the repetition may be exact (in which case repetition-
marks :|| will be used, possibly with ist and 2nd ending), or it
may be unessentially modified.

The questions : Whether to repeat or not ? Whether to repeat exactly
or not ? and, if the latter, How little or how much unessential variation to
introduce ? are touched upon in par. 23, which review. See also par. 103.
From these, the general principle will be inferred that the advisability of
repetition depends: (i) upon the simple vequirevaent oi dimension or propor-
tion, and (2) upon the necessity of emphasizing or confirming a somewhat
complex^ or a sufficiently attractive^ sentence. Furthermore, a simple sentence
will call for more elaborate alteration than a complex one.

k

Par. 106. ELABORATE REPRODUCTIONS OF SECOND DIVISION. I97

(b) The rule against the repetition of the Second Part alone,
while quite strict, is not infrequently disregarded when the Part
is small (as in Ex. 78 and Ex. 91) ; or when the repetition is
followed by additional retransitional material (as in Ex. 84) ; or
when the repetition, for any other evident reason, does not tend
to sever the connection with Part III. Other, distinctly irregular,
cases of the repetition of Part II alone, are shown in par. 11 2^.

(i) For illustrations of exact rep, of Part /, see Exs. 81, 84, 85, 89;
Mendelssohn, ** Songs without Words," Nos. i, 3, etc. ; Grieg, Lyric Pieces,
op. 43, No. 5.

(2) An exact repetition oi Parts II and III taike^ place in Mendelssohn,
** Songs without Words," Nos. 12, and 45 ; Grieg, Lyric Pieces, op. 12, No. 5,
No. 7; op. 38, No. 3, No. 6; op. 43, No. 3 (final extension added).

(3) Both Divisions are literally repeated in Beethoven, Pfte. Son. op. 2,
No. I, both Menuetto and its ** Trio " ; Mendelssohn, ** Songs without
Words," No. 7; Grieg, Lyric Pieces, op. 12, No. 4; op. 38, No. 4.

(4) For illustrations of the repetition of Part I with unessential modifica'
iions, see Mendelssohn, ** Songs without Words," Nos. 4, 9, 16, 21, 27, 31,
46; Beethoven, Pfte. Son. op. 26, Scherzo; Schubert, Momens musicals
op. 94, No. 6, ** Trio " ; Chopin, Mazurka No. 37 ; Haydn, Pfte. Son. No. 15
(Cotta ed.), Adagio ; Brahms, op. 118, No. 2, first 48 measures; op. 118,
No. 6.

(5) Modified repetition of the second Division (Parts II and III) will be
found in Mendelssohn, "Songs without Words" Nos. 19, 23, 48, 8; Beet-
hoven, Pfte. Son. op. 53, Finale (Rondo) first 62 measures; Schubert,
Momens musicals op. 94, No. 2; Chopin, Nocturne No. 11 (op. 37, No. i)
Principal Song; Nocturne 13, second tempo (C major); Polonaise No. 5 (op.
44) first 78 measures (introd. 8 meas.) ; Schumann, op. 99, No. 5 {b minor).

(6) Finally, both Divisions are repeated with modifications , in Mendels-
sohn, "Song without Words" No. 29; Haydn, Pfte. Son. No. 14 (Cotta ed.)
Adagio^ first 56 measures; Chopin, Nocturne 2 (op. 9, No. 2); Nocturne 3,
first tempo; Nocturne 15 (op. 55, No. i), first 48 measures; Mozart, Pfte.
Son. No. II (Cotta ed.) Andante,

EXERCISE 32.

An example of the 3-Part Song-form, each Division (Part I alone, — Parts
II and III together) repeated^ with unessential variations. See par. 130, which
may be applied here, if deemed desirable. See par. 137.

2. More elaborate reproductions of the second Division.

1 06. The entire line of development, beginning with the
simple tripartite form and leading up to the genuine 5-Part
Song-form, does not, in any sense, concern the manipulation of

198

BI^ABORATS REPRODUCTIONS OF SECOND DIVISION. Par. 106b.

the First Part ; but refers exclusively to the reproduction of Parts
JI and III. It runs through five successive and clearly distin-
guishable Stages, definable as follows, according to the method of
treatment adopted in reproducing the Second Part :

(a) The 1ST Stage in the evolution of the 5-Part form is
represented by the exact, or unessentially modified, repetition
of the second Division (irrespective of the treatment of Part I), as
indicated in par. 105, and illustrated in the fifth set of references
given above. As nothing beyond the idea of simple Repetition is
herein embodied, and as ** repetition effectuates no actual advance
in structural design," whether modified {unessentially^ or not, —
see par. 21 ^ last clause, — the term ** 3-Part " form must be adhered
to in this ist Stage.

(b) In Stage 2, a more significant alteration of the original
Second Part is made (upon the reproduction of the latter as Part
**Four"), consisting in its recurrence in a different key. Here,
already, the term " Five-Part " Song-form may be adopted.

This transposition of the Second Part may be partial, or
complete. In the latter case, the Part is transferred bodily, with
little or no further change, a certain interval upward or downward ;
usually, though not necessarily, into some closely related key.

There will be no difficulty in making the modulation into the new plane,
because of the (presumably) complete cadence at the end of the Third Part, —
which the transposed version of Part II is to follow. And the return to the
original key can easily be accomplished during the usual, — ^possibly special, —
Retransition. Par. 90c. For illustratiQn :

Presto,

100.5

Part I. (Period, 8 measures.)

izfcfzi'zt:^

I

T— etc. ; «ee—
Original. •!)

Part XL

V

ur^\ i n i h

v^^

Far. 108b. ELABORATE REPRODUCTIONS OF SECOND DIVISION,
Exteusioii y

199

g

31^^

t ♦

-# — 0-

=#:

S

K ' " -^

:i'e±/'"A'^^

T T J f ; J /L i i. j

j j ^ —r

Ketraiisitiou.

Part III. (GroupofaPhnwes).

^te

• •

i

etc.; see Original.

Part IT. (transposed recurrence of Part H, complete). *2)

5e

^^nm^t^i

A Minor.

4

^

^■a- i ffl-i^ i ^

^..''.j'' PI

Extension.

?^

V

^€' *

^m

±±

EI

^^S

*^

Eetransition.

^ ^ ^-^^' i^ i ^ ^i^T^-j^^j^-

*3)

%^ Mf^^^g^''^{t^x;'^;'^ ' »^ ^

PartV.

i

(recnrrenoe of IH ).

Mrndblssohn.

■+■

H:

^^

5^

i

pp etc.; see Original.

200 ELABORATE REPRODUCTIONS OF SECOND DIVISION. Par. 106c

♦i) Mendelssohn, ** Song without Words" No. 34.

*2) The entire Second Part is transferred bodily downward a 3rd ; and
the transposition is literal, only excepting in measures 8 to 10 (chiefly during
the ** extension'*).

*3) At this point, the purpose — and the method — of regaining the original
modulatory line, are displayed. The Fourth Part ends upon E^ a 3rd lower
than before; here the same Retransition is again utilized, but it gradually
shifts upward, in admirable keeping with the nature of the figure, until, at
Note *4), the original location, upon G, is reached.

See Chopin, Mazurka No. 33 ; the transposition is not literal, but constant.

Rubinstein, Le Bal, op. 14, "Contredanse,** No. ^{Alleg-retto, % time);
transposition not literal, but aflPecting the entire Part.

Schumann, Bunte Blatter (op. 99) No. 12 (** Abendmusik"), Principal
Song; Part II distinct and repeated; Part III also repeated, alone.

If, as is somewhat more likely to be the case, the transposition
of Part II (as Part IV) is not to extend uniformly through the
entire Part, then the melodic line of Part IV may either be so
conducted as to return at some point (gradually or abruptly,
as proves most convenient) to the original line of Part II; or
Part IV may be abbreviated, or otherwise so manipulated as to lead
properly into the key of Part V.

See Mendelssohn, ** Song without Words" No. 24; Part IV sets out
a step higher than Part II ; the transposition is not literal, but constant, so
that the Parts are of exactly similar length; in meas.*i3 of Part IV, the
original melodic line is regained, then again abandoned for \^ measures, and
then so guided as to close precisely as Part II did.

In **Song without Words" No. 14, Part IV begins a fifth lower than
Part II does, and adheres to this interval of transposition up to the 9th
measure, whereupon a skip of four measures occurs, leading abruptly back
into the original melodic line of Part II, the last 3 measures of which recur
literally.

In No. 44, the transposition is only partial, affecting the second half {two
measures) of Part IV, which is a 3rd higher than the corresponding measures
of Part II ; the change of key spreads slightly beyond the limit of the Part,
however, and involves the first chord of Part V. (See par. no.)

(c) In the 3RD Stage, the idea of ** transposition " still

prevails, for a longer or shorter period, at the beginning of Part

IV ; but the change of key gives rise, naturally, to other ^ more

radical, changes in the thematic structure of the Part^ which

sometimes assume significant proportions.

See Mendelssohn, op. 72, No. 5 ; Part IV begins a step higher than the
Second Part does, and pursues the melodic line of the latter quite closely for
5 measures; then digresses, and expands, but closes just as Part II did. —
"Song without Words" No. 43; Part IV lies a 4th above the Second Part

Par. 106€. ELABORATE REPRODUCTIONS OF SECOND DIVISION. 20X

for 4 measures; then diverges, adds 3 **new'* measures, and leads into an
intimation of the First Part (as Part V). This intimation is almost vague
enough to justify calling these two ** Parts " a Coda. — ** Song without Words'"
No. 17; Part I is a small Period, 4 meas., closing with an imperfect cadence;
Part III has exactly the same melody, but modulates into a similar imperfect
cadence, thus inaugurating the transposed recurrence of Part IV, as a natural
consequence; Part IV utilizes two measures of Part II, then diverges, extend-
ing the given material in an independent manner, and, omitting all the rest of
Part II, closes in the 7th measure; a long Retrans. follows.

Cramer, Etude 17 (orig. compl. edition) ; Part IV is much longer than
Part II, and consists of two similar Sections, both ending with the same
complete cadence. This last example might, perhaps more accurately, be
classed among those of the 4th Stage.

Chopin, Nocturne 8 (op. 27, No. 2); Part I 9 measures, imperfect cadence ;
Part IV follows the melodic line of Part II quite closely for 8 measures (at first
a 6th above, then a 2nd below), and then leads back to the beginning with
a shorter and more emphatic Retransition.

Schubert, Pfte. Son. No. 10 (B(> major). Finale, first 73 measures; Part I
repeated ; Part IV begins exactly as Part II does, but soon diverges, trans-
poses, extends and adds new material, to considerable length.

(d) In the 4TH Stage in the evolution of the 5-Part Song-
form, the Fourth Part no longer adheres strictly to any portion of
the melodic line of Part II (in the same key, or a different one), but
is a completely *' reconstructed'''^ recurrence of the latter. This
reconstruction, however, is conducted along lines corresponding
generally to the former contents and style of Part II, so that a more
or less palpable resemblance between the two versions is preserved.

This is admirably illustrated in Chopin, Prelude op. 28^ No. 17; Part II
begins, in meas. 19, with an enharmonic change ; Part III contains only the
first half of Part I, and ends with the semicadence at that point; Part IV also
starts out with an enharmonic change, imitates the conduct of the Second Part
for a while, but is longer and more elaborate than the latter. Observe, par-
ticularly, the dynamic design, also, of this composition (par. 95«).

See further, Mendelssohn, Praeludium op. 35, No. 2 (D major); Part IV
is very much longer than Part II, and resembles the latter most closely in its
later course. — Schumann, Arabesque (op. 18), **Minore I.^'

(e) In Stage 5, finally, all palpable resemblance disappears,
and Part IV becomes a distinct and individual member of the
form, — independent of the Second Part, notwithstanding their
coincidence of purpose. For this, the fully developed, 5"^^^
Form, a new title might be adopted, indicative of its characteristic
and inviolable derivation from the tripartite design, namely : A

202 TREATMENT OF PART FIVE. Par. lOT.

3-Part form with two different '' second " (or ** middle ") Parts, or
** Departures " (par. 8ia). It is the type of the Form explained

in par. 124.

See Schumann, Kreisleriana (op. 16) No. 6 ; Humoreske (op. 20), ** Einfach
and zart " ; same work, ** Innig " ; Nachtstucke (op. 23), No. 4.

Beethoven, Pfte. Son. op. 13, Adagio; Parts I and V repeated. — Mozart,
'* Fantasie and Sonata" in c minor, Adagio of the Sonata (Cotta ed. No. 18).
These two examples are both unusually broad in design, and approach a certain
grade of the Higher Forms, in connection with which they will again be cited,
m another volume.

Chopin, Mazurkas, Nos. i, 2, 5, 8, 15, 30 (long Coda), 48. — Pfte. Son. No. 2
(op. 35), ** Pid lento *' of Scherzo (Parts IV and V repeated).

Rubinstein, Le Bai{pp. 14), ** Contredanse," Nos. i, 2, 4.

Treatment of Part Five.

1 OT- In all of the above cases of modified reproduction of
the second Division, the Fifth Part is an additional '''' return'*'' to
the beginning (or da capd)^ and therefore it is likely to corroborate
Part III. But it is not necessary that it should correspond exactly
to the latter. All the various modes of treating the Third Part, —
the several stages of divergence from, and elaboration of, the First
Part, enumerated in par. 102a, 3, r, and d (and par. 99^, 3, c),
— may be exhibited in Part V, irrespective of the species of
manipulation previously chosen for the Third Part. Hence it is
possible that Parts I, III and V, while essentially corresponding
in contents, may all differ in treatment, each representing a com-
paratively independent version of the original ** Statement" (par.
8itz) ; perhaps in progressive ratio, — ^Part V becoming the longer
and more elaborate of the three versions (comp. par. 53) ; but, on
the other hand, it is equally consistent, though less usual, to
abbreviate the third '* return," especially in lengthy compositions,
for obvious reasons.

In Mendelssohn, ** Songs without Words," Nos. 19 and 24, Part V is
exactly like Part III ; in No. 48 it is a slightly variated recurrence; in Nos. 34
and 23 it is extended a trifle at the end ; in No. 14 it is considerably changed
and extended ; in No. 8 it is enriched by the addition of new material.

In Mendelssohn, Praeludium, op. 35, No. 2, Part V reproduces a larger
portion of the First Part than the Third Part does (which is unusually brief),
then adopts the whole of Part II, and then adds new material, — but without
disturbing its own continuity.

In Schumann, Kreisleriana (op. 16) No. 6, Part III is an extended version
of Part I, while the Fifth Part returns to the original form and reproduces
Part I almost literally.

Par. 108b. OLD-FASHIONED RONDEAU, AND 7-PART FORM. 203

In Schubert, Pfte. Son. No. 10 (B|^ major), Andante Movement, Subordi-
nate Theme (3-sharp signature), the First Part is repeated with a characteristic
change of style; after Part III is closed, a similar repetition of the second
Division is started^ with the same change of style, and with every indication
of genuine purpose ; but upon approaching the end of Part II (IV) the purpose
is abandoned, for reasons incidental to the Higher Form of which it is a portion,
and, consequently, Part V does not appear at all.

In Schumann, op. 82, Nos. 6 and 9, a somewhat similar omission of
Part V (the final da capo) occurs, though here it is less convincing, because,
unlike the Schubert example. Part IV differs from Part II. These examples
will be cited again, in par. 115.

The Old-fashioned Rondeau, and the 7-Part form.

1 bSa. The 4th and 5th Stages of the Five-Part Song-form,
explained in par. 106^ and e^ correspond in design to the most
common species of the ** Rondeau," — a form that was in vogue in
the 17th and i8th centuries, and subsequently developed into the
more elaborate modern ** Rondo."

Illustrations may be found in "Les Maltres du Clavecin*' (Litolff ed.).
Vol. II, pages 142, 148, 158, Rondeaus by J. Ph. Rambau.

(b) The fundamental idea that has been seen to result in the
enlargement or growth of the tripartite design into one of five
Parts, is, plainly, that of multiplying the ''^Departures'^'' from
the original Statement (Part I), in connection with each ensuing
*' Return." This idea was quite frequently carried still farther, by
older writers ; more, — it must be observed, — with an eye to mere
length, than to compactness and true beauty of form. Thus, the
old-fashioned Rondeau often embraced three and even more differ-
ent, but kindred, ** Departures " (or *' Counter-themes "), each in
One-Part form, followed by as many recurrences of the First Part
('* da capos ").

See J. S. Bach, Partita II, "Rondeau" (three departures); J. Ph.
Rameau, "Les Mattres du Clavecin" Vol. II, p. 144 (three departures);
Francois Couperin, same publication. Vol. II, pages 116, 126 (three depar-
tures).

In its application by later, and by some modern, composers
to the system of musical designs, it has given rise to a form which,
by analogy with the above, must be denominated the '* Seven-
Part Song-form." Fortunately for the purity, concentration
and stability of musical architecture, these attenuated, restlessly

204 THE IRREGULAR PART-FORMS. Par. 105>.

revolving structural designs, whose only logical justification is that
of alternation, are very rarely adopted. -The student of musical
Form should shun their imitation.

See Schumann, NachtstUche (op. 23), No. i; 7-Part Song-form (i. e. a
principal Part, — parallel Period, — with 3 different and distinct ** departures,"
and as many da capps\ followed by a long Coda in which even Parts VIII and
IX are intimated. — Same opus (23), No. 2, measures 25 to 51 ; 7-Part form on
a diminutive scale (Part I, 4 measures ; II, 2 measures ; III, 4 measures ; IV= II
transposed; V, 6 measures; VI, 5 measures; VII, 4 measures, followed by
Retransition into the Principal Song). — Schumann, op. 15, No. 11; 7-Part
form, the 3rd departure (Part VI) identical with the first one (Part II).

EXERCISE 33.

A. An example of the 5-Part Song-form, Stage 2 (par. io65) ; all detailed
conditions optional. Reference may be made to par. 132. See par. 107.

The pupil should, hereafter, endeavor to sketch his work rapidly,
especially the principal melodic line or lines, — unless his musical
disposition be such as to render such a desirable process impracticable or
insurmountably difficult. In any event, it is wise to fill out as much of the
details as possible with all necessary accuracy at leisurcy after the main design
is fixed. See par. 137.

B. An example of Stage S (par. io6c). See par. 107. — Application of
par. 129, or 130, may be made, if desirable.

C. An example of Stage 4 (par. io6e/). See par. 107. — Application of par.
133 may be made.

D. An example of Staged (par. io6e). Review pars. 87, 88, 89, 90; pars.
95, 96; par. 107. — Application of par. 134, or 135, may be made.

CHAPTER XVII.

THE IRREGULAR PART-FORMS.

1 09, There are a few other varieties of these musical designs
(consisting in the association or compounding of ' ' Parts ' ' ) which
violate one or another of the essential conditions of the regular
structural plan, and must, therefore, while being accepted and
sanctioned, be qualified as *' irregular." The pupil can not afford
to remain in ignorance of them, but must estimate them as very
rare, and abstain from their too frequent use.

par. 110. THE TRANSPOSED THIRD PART. 205

I. The transposed Third Part.

* 1 10. K da capo^ or recurrence of the First Part, beginning
in any other than the principal key, is naturally hazardous, because
the resumption of the original key (after probable absence from it
during Part II) is quite as vital an indication of the Third Part, as
is the recurrence of thejirst melodic member. But a harmless trans-
position is nevertheless conceivable under the following conditions :
(i) That the beginning of Part I is of so striking a character
in melodic, rhythmic or harmonic respect, that its recurrence is
sufficiently convincing, despite the change of key ;

(2) That Part III appears just where it is expected ;

(3) That it appears at nearly, or quite, its full length ; and

(4) That a decided return to the principal key is effectuated
during the course of the Third Part.

•

Further, the end of Part II should be plainly recognizable.

The transposition may extend through the entire Third Part,
or through only a portion of it ; but will always, in this phase of
irregular form, affect its beginning.

Illustrations of such "tampering with the beginning" of Part III as
paves the way to the deliberate transposition of a section, or all, of the Part,
have been cited in the notes to Ex. 88, and context, and are repeated here for
renewed reference: Mendelssohn, ** Songs without Words," Nos. 40, 41, 36,
37 (first measure of Part III) ; Chopin, Mazurka 46 (meas. 24-25) ; Nocturne 16
(op. 55, No. 2, meas. 9) ; Schumann, Waldscenen (op. 82) No. 4, — d minor,
meas. 23 (comp. with meas. i); Brahms, Balladen op. 10, No. 4, meas. 27-28
(comp. with meas. 1-2) ; Pfte. Pieces, op. 117, No. 2, meas. 51-54 (comp. with
meas. 1-3) ; op. 118, No. 2, meas. 34-36 (comp. with meas. 1-2, of which they
are the "contrary motion"); op. 118, No. 6, meas. 59-62 (comp. with meas.
1-4) ; op. 119, No. 2, second tempo (4-sharp signature), measures 25-26 (comp.
with meas. 1-2); op. 119, No. 3, measures 41-42 (comp. with first measure, of
which they are an ** augmentation ").

In some cases the transposition of Part III does not affect
the original Tonic^ but simply changes the mode from major to
minor, or vice versa. But usually some other (almost invariably
next-related) Tonic is chosen ; most frequently the Subdominant.
These various conditions, concerning the manner, the extent, and
the modulatory relation of the transposed Third Part, are illustrated
in the following :

Schumann, Bunte Blatter (op. 99), No. 11, — March in ^/-minor, meas.
27-30 (a third higher than first Phrase of Part I); same work, No. 3 (Part III
in Subdom. key) ; Arabesque^ op. 18, ** Minore II ** (Part II very independent).

206 THE GROUP OF PARTS, INCIPIENT STAGE. Par. 111.

Brahms, op. Ii6, No. 5; (Part II ends on Dom., as usual; Part III begins
in Subdom. key, and follows the First Part closely for four measures, then
becoming more independent).

Grieg, op. 38, No. 8, first tempo.

Beethoven, Symphony No. 3, Funeral Marchy first Theme; (Part III
begins in Subdom. key; Parts II and III repeated; Codettas). — Pfte. Sonata^
op. 14, No. I, Finale^ one-sharp signature. Pfte. Sonata, op. 22, Finale y
measures 72-103.

Bach, Well-tempered Clavichord, Vol. I, Prelude 9; Vol. II, Prelude
19 (in each case. Part III begins in the Subdominant key) ; Vol. II, Prelude
22 (the same ; Part II begins in meas. 16, Part III in meas. 55).

Schubert, Momens musicals op. 94, No. i, ** Trio" (one-sharp signature;
Part III the Opposite mode of Part I for 4 measures) ; Impromptu op. 90*
No. 4, " Trio" (four-sharp signature; similar); Fantasie in G^ op. 78, " Trio "
of Menuetto; String-quartet in Ay op. 29, Menuetto.

Chopin, Nocturne 12 (op. 37, No. 2), meas. 30-69. — Impromptu, op. 36
(/^-major) ; a unique example, the Parts very broad, very elaborate Coda.

Sometimes this transposed recurrence is applied to the Fifth
Part (par. 107).

And it is not infrequently involved by some modulatory design
extending through an entire 5-Part Form.

Haydn, Symphony 12 (Peters ed.). Adagio; (Part V transposed, extended,
and reproduced in original key ; quasi 7-Part form, with Coda).

Chopin, £tude op. 25, No. 3 (Parts III and IV transposed ; Part V again
in original key ; quaint modulatory design).

Schumann, Faschingsschwank op. 26, movement IV, *' Intermezzo**
(key-scheme : Part I, ^-^~F\ Part III, ^-a\)-B^ ; Part V, d^-^^-E^). These
last two unique examples should be carefully examined.

2. The Group of Parts, Incipient Stage.

111. The distinction made in chap. VII (particularly pars.
57, 58, 59) between the regular, coherent forms, and the more
loosely connected group-formations, may be extended to the
Part-forms also. Thus, the ** Parts," even in the 3-Part form,
are sometimes so loosely associated^ with so little regard to the
principles of logical continuity and cohesion, that only the term
** Group" may justly be applied. This disintegration of the form
is initiated by all such liberties in the structure of the Second Part
as tend to isolate it from its fellows, either in consequence of
thematic independence and diversity, or of more qr less complete
separation by a full cadence.

Par. 118. THE GROUP OF PARTS, INCIPIENT STAGE. 207

112. Some of these irregularities have already been inti-
mated — in par. 87^, and 105^; but it is necessary, here, to
enumerate them more specifically.

(a) Cases of undue isolation or distinctness of Part II,
caused by striking difference in style or thematic structure, or by
complete detachment (at the cadence) from Part III, are exhibited
in the following :

Schumann, Kreisleriana (op. 16) No. 4; Albumbldtter (op. 124) No. 3.

Chopin, Nocturne 4 (op. 15, No. i). — Brahms, op. 118, No. 5.

In Brahms, op. 117, No. 2, the Second Part is thematically almost
identical with Part I, but radically different in style, through its entire length.

In Brahms, op. 116, No. 2, and Grieg, op. 38, No. i, the distinct Second
Part is followed by quite an elaborate Retransition.

(b) A REPETITION OF THE SeCOND PaRT ALONE often OCCUrS

in direct consequence of such distinctness of character. Compare
par. 103 ; and par. 105^5, last clause. For illustration, see : ^

Beethoven, Bagatelle op. 33, No. 2; Pfte. Son. op. 14, No. 2, Finale^
meas. 73-124 (Part II repeated and extended ; followed by a Retransition). —
Ex. 91 of this book. — Chopin, Mazurka No. 24. — Schumann, op. 124, No. 4;
op. 15, No. 6. — Grieg, op. 38, No. 7. .

An extraordinary example is found in Beethoven, Pfte. Son. op. 90,
Finale^ first 32 measures; peculiar, not only because of the (apparently
uncalled-for) repetition of Part II, but chiefly for its modulatory design; each
of the three Parts, — including Part II and its repetition^ — closes with a perf.
cadence in the principal key. Such daring experiments as this are for the
pupil's amazement, not for his imitation; he must defer them to the period
of his most absolute maturity. — Compare par. 88.

113. Indirectly analogous to the Transposed Third Part
(par. no), is the skqjljkntial, or transposed, reproduction
of one or another of the Parts, — a process which also militates
against the stability of the structure. See :

Bach, Well-tempered Clavichord, Vol. II, Prelude No. 11 (F major);
modified sequence of Part I, in meas. 17-32.

Chopin, Prel. op. 28, No. 24; sequential reproduction of First Part, in
meas. 21-38. Nocturne 18 (op. 62, No. 2), meas. 32-57 ; Two-Part form, Part
II sequentially reproduced (a third higher). Nocturne 10 (op. 32, No. 2) piii
agitato ; Two-Part form, reproduced bodily a half -step higher.

Schubert, Fantasie in G (op. 78) Andante, meas. 31-49 ; a 2-Part form with
Codetta, followed, in the next 19 measures, by a slightly modified sequence.

Brahms, Pfte. Son. op. i, Finale^ second Subject (one-sharp signature) ;
Third Part extended by elaborate transposed reproduction.

2o8 GROUP OF FARTS, DEVELOPED, AND EXTENDED. Par. 115.

3. Group of Parts, developed, and extended.

114. The disintegration of the regular Song-form, of which
the above paragraphs exhibit the incipient stages, becomes complete
when Part number TTiree (unmistakably identified as such by
its style, and by the foregoing cadence) does not in any sense
correspond to the First Part, Comp. par. 8ia. For such a series
of three independent Parts, no other epithet than *' Group" can
properly be employed. Review 69^, and 70. See :

Haydn, Symphony No. 9 (Peters ed.), Finale^ principal Theme. (Part I,
8-meas. Period; Part II, 8-meas. Period, with strong perf. cadence. on original
Tonic ; Part III, i6-measure Double period, — suggestive of the foregoing Parts,
but unquestionably independent of them) .

Schumann, Papillons (op. 2) No. 9, — Parts II and III similar, each Part
repeated.

115. Such examples of the Group-form as are thus limited
to three Parts, are, however, very rare. It is far more usual, when
this irregular design is adopted, to extend it to four or more Parts.

And, quite frequently, the specific principle of the tripartite
forms is manifested by a ' * return to the beginning ' ' (Part I) , in
one of the later Parts, — generally the last one. In this case, espe-
cially, the possibility, and occasional presence, of an independent
Coda to the Group may be demonstrated. For illustrations, see :

Beethoven, Bagatelle op. 119, No. 3 (five Parts, each repeated, and
Coda; Part V like I). — Schubert, Momens musicals op. 94, No. 3 (four Parts
and Coda; Part IV like I).

Chopin, Nocturne 5 (op. 15, No. 2; — ^four Parts, and Codetta; Part IV
like I). Nocturne 6, — four distinct Parts. Nocturne 9 (op. 32, No. i ; peculiar ;
Parts I to III regular, — III like I ; followed by a Fourth Part which embraces
a transposed recurrence of Part II, and is repeated, with Codetta-extension).
— Mazurkas, Nos. 3, 7, 20 (five Parts, V like I); Mazurkas, Nos. 14, 19, 21, 39
and 41 (four Parts, IV like I); Mazurka No. 27 (five Parts, thus: I, II, III, II, I);
Mazurka 34 (six Parts, VI like II ; each Part repeated) ; Mazurka 35 (seven
Parts; I repeated; VI like III; VII like I; VII reproduced, "dissolved" and
extended; Coda).

Schumann, Waldscenen (op. 82), No. 6, represents an ostensible 5-Part
form (Part III like I) with omission of the final da capo (Part V). Such
deficiencies are usually made partly good in the Coda ; but in this case there
is no more than a trace (in the last measure) of such a compensation. Op. 82,
No. 9, is almost precisely the same. In the Papillons (op. 2), No. 8, the form
is ostensibly 3-Part, with introductory Part (four in all).

Par. 115. GROUP OF PARTS, DEVELOPED, AND EXTENDED. 209

Brahms, Capriccioy op. 76, No. 5 (six Parts, III like I, V like II, VI
like III, brief Coda; the form is quasi 5-Part, first stage, — par. io6«, — ^with
interpolated Part after Part III) .

BXBRCiSB 34.

A. An example of the 3-Part Song-form with transposed Third Part
(par. no). Choice of key optional ; and the transposition may extend through
the whole, or only a portion, of Part I. Apply par. 129, or 130. See par. 137.

B. One example of the Group of Parts, according to par. 115. Apply
par. 133, or 131.

ariO THE SONG-FORM WITH ONE "TRIO. Par. 117.

DIVISION THREE.

THE COMPOUND SONG-FORMS.

116- The last degree of enlargement and development
possible within the domain of the homophonic forms, is achieved
by associating' complete Song-forms, See par. 69c.

The association is effected in general accordance with the rules
governing the union of Parts in the 3-Part and 5-Part forms, but
with less stringency. Thus, while the law of '* Recurrence " must
be respected, more latitude is permitted in regard to thematic and
formative relation between the several Song-forms; and uninter-
rupted connection at the points of contact, is the exception, rather
than the rule.

CHAPTER XVIII.
THE SONG-FORM WITH ONE "TRIO."

1 IT. The most common of these Compound forms consists
in the association of tivo Song-forms. The one which comes first
is called the Principal Song (chiefly by virtue of its location, and
its recurrence) ; the second one in the series is commonly called the
**Trio" (for reasons derived from a now obsolete custom), but the
designation adopted in this book, in analogy with the terminology
of all Higher forms, is *' Subordinate Song."

After the latter, the Prin. Song recurs, as a '*Da Capo,"
conformably with the ruling principle of all tripartite designs.
Thus, the Song-form with one * ' Trio ' ' is perceived to be a broader
exposition of the 3-Part Song-form, in which each *'Part" has
expanded into a complete ** Song-form." Compare par. 77;
8ia; 91.

Besides the popular term "Trio," many others are employed by different
composers to indicate the Subord. Song in their Compound Song-forms;
e.g., ** Alternativo," "Intermezzo," "Musette" (in the Gavotte), "Fris"
or "Friska" (after the "Lassen" in the Hungarian Czardas). In older
compositions, the successive Song-forms were sometimes simply numbered:

Par. 118c. THE PRINCIPAL SONG. 211

Menuet I, Menuet II ; Passepied I, Passepied II, etc. (Bach, 2nd English
Suite). Sometimes the terms ** Minore " and ** Maggiore *' are used, indicating
at the same time the corresponding change of mode for the Subordinate Song ;
and frequently, for a similar reason, the sign of altered tempo or character
(**pivi lento," **meno mosso," etc.) serves to denote the second Song-form.
Finally, all external indication is often omitted, save perhaps a change* of
signature. The derivation of the term **Trio" is illustrated in Bach, 3rd
French Suite, Menuet (Prin. Song for two melodic parts, Subord. Song for
ilireey hence a ** Trio *0.

The details of the Song-form with '*Trio" are as follows:

The Principal Song.

1 1 8a- The Principal Song may be of any character, and
is constructed most frequently in the 3-Part Song- form, with all
repetitions ; possibly with a brief Codetta. As a rule, it closes with
a strong perfect cadence in its own key, sufficient to constitute
absolute independence of form.

See: BEETHOVEN<Pfte. Son. op. 2, No. i, MenuettOj first Song (40 meas-
ures) ; 3-Part form, all repetitions. — Pfte. Son. op. 2, No. 2, Scherzo (ditto). —
In Haydn, Pfte. Son. No. 6(Cotta ed.), first movement, the Principal Song
is repeated entire, as Variation, before the Subord. Song appears.

(b) A Prin. Song in 2-Part Song-form is very unusual; and

still more rarely is the One-Part form chosen. In the latter case,

the Subord, Song must be at least 2- or 3-Part form, and very

distinct in character.

See : Chopin, Mazurka No. 50, first 32 measures (2-Part form, each Part
separately repeated). — Beethoven, Pfte. Son. op. 31, No. 3, Menuetto (ditto);
Pfte. Son. op. no, second movement, first signature (2-Part form). — Brahms,
op. 116, No. 3, first 34 meas. (Prin. Song only One-Part form, — ext. Double
period ; but the following Subord. Song is 3-Part form and distinct). Op. 116,
No. 7 (similar). — Hungarian Dances, No. i (Prin. Song 2-Part form, each
Part repeated)^ Nos. 5, 6, 9, similar. — Chopin, Nocturne No. i (Prin. Song,
One-Part).

(c) In some comparatively rare cases, a very brief Transition
intervenes between the Prin. Song and the following '*Trio." It
serves as a mediating link between the different styles and keys
of the two Songs, and becomes more necessary, and more extended,
in proportion to the degree of their differentiation. It need not
impair the independence of the Prin. Song, for the complete
cadence of the latter* may precede the transitional passage ; though
sometimes this cadence is so modified or concealed that the Tran-
sition may emerge from it.

212 THE SUBORDINATE SONG, OR **TRIO." Par. llflL

See: Beethoven, Pfte. Son. op. 14, No. i, third movement (Prin. Song,
3-Part form with Codetta ; one measure of Transition). — Haydn, Symphony
No. 3 (Peters ed.), Menuetto (2 meas. of Transition precede the "Trio"). —
Chopin, Prelude op. 28, No. 15 (Prin. Song 3- Part form, cadence evaded —
measure 27 — by resting on Dominant). — Schumann, Waldscenen (op. 82), No.
8^ Prin. Song 3-Part form, long Codetta; the cadence-measure transitionally
bridged.

The Subordinate Song, or **Trio."

1 19. The Subordinate Song, or "Trio," should contrast
quite positively in general character with its Principal Song,
though radical or extreme difference of style, to the utter exclusion
of organic interdependence, consistency, and congruity, must be
avoided. In some examples, especially older Dances, there is
scarcely any recognizable diversity between the two Songs ; but
in more modern composition, contrast and separation are the laws
of the Subord. Song.

In Beethoven, Pfte. Son. op. 2, No. i, Menuetto^ the Prin. Song is
rhythmically marked, and somewhat dramatic; the '*Trio" is smooth and
more lyric ; but the organic harmony between them is perfect- In the next
Sonata, op. 2, No. 2, Scherzo^ the contrast is more striking; and in the 3rd
movement of the next (op. 2, No. 3) it is still more emphatic; but in both,
admirable consistency is preserved.

Thematic relation of the Subord. Song to its Prin. Song is
"uncommon, ^nd, when it appears, it must be counteracted by more
complete contrast in style.

See: Haydn, Pfte. Son. No. 6 (Cotta ed.), first movement; the first figure
of the Subord. Song corresponds closely to that of the Prin. Song; the rest
is thematically independent, but similar in style. — In the next Son. (No. 7),
Finale^ the Prin. Song is in 2-Part form; the Subord. Song resembles it, for
a time, so closely that it appears to be no more than a Variation of the former ;
but it is in 3-Part form, and in the opposite mode. — In Brahms, op. 116, No. 3,
and op. 119, No. 2, the thematic material of the two Songs is identical, but the
change in style almost radical, especially in the latter example.

But the entire perplexing question of analogv(consistency,
inner and outer harmony) between the two Songs in this Compound
form, depends more largely upon the correlation of Time, K ey an <
Tenipo. These establish the requisite externaT^xdoxva\\.y without
efrcfoaching upon the conditions of contrast and separation ; while
thematic coincidence may only ensure the more vague and (in
homophonic forms) comparatively unessential inner affinity.

Par. 119b. THE SUBORDINATE SONG, OR "TRIO." 213

(a) The Time of the Subord. Song is almost invariably the
same as that of its Prin. Song; i. e., both are either in duple, or
triple, measure. Exceptions :

In Chopin, Gl>-major Impromptu, and Nocturne No. 10, C or $ and V
time are interchanged. In Nocturne 14 (op. 48, No. 2) the Prin. Song is in
\ time, and the **Trio" {molto piU lento) in J. — In Brahms, op. 116, No. 7, the
time changes from | to j. — Beethoven, Symphony No. 6, third movement;
change from } to }. — It must be emphasized, that this is exceedingly un-
common.

(bX J^e Key of the Subord. Song is always related, in some

degree, to that of the Prin. Song, though unlimited option prevails
in the choice of relationship (near or remote). In the oldest, and
in some modern, examples, the selfsame key is retained ; in which
case the separation of the Songs, and their respective completeness
of form, is more strongly marked :

Haydn, Symphonies (Peters ed.), Nos. i, 4, 5, 6, 7, 8, 10, 11, 12, — third
movement of each. Beethoven, Symphonies, Nos. i, 2, 3, 4, 6, 8, — third
movement of each. "^ ■• ""^

The next step to this, in the early history of the **Trio,'*
"appears to have been the choice of the Opposite Mode of the same
Tonic :

BE gyHOVEN, Symphony No. 5, third movement (c-minor and C-major);
Pfte. Sonatas, op. 2, No. i ; op.^'No.'S'; op. 7, — third movement of each.
Brahms, op. 119, No. 2. — Chopin, Polonaise No. i.

The Relative (major or minor) key is found in :

Beethoven, Pfte. Sonatas, op. 2, No. 3, third movement (C-major and
a-minor) ; op. 22, op. 28, third movement of each.

Next to these in popularity is the choice of the Suhdominant
key (or its Relative) for the '*Trio"; this, in the modern March
and in many Dances, has become a usage almost equivalent to
a rule. It is notably appropriate in the homophonic domain of
musical architecture, because the inclination to relax toward the
lower (Subdom.) keynotes, — in distinction to the aspiring impres-
sion conveyed by the upper (Don^nant) ones, — is characteristic of
the inferior range of forms. This point, which is of vital moment
in the Complex forms, will be reverted to in connection with the
Rondo- and Sonata-forms, in a subsequent Volume.

/

214 THE SUBORDINATE SONG, OR *'TRI0." Par. llQd.

The Subdom. key appears in Beethoven, Pfte. Son. op. lo, No. 3, third
movement, (Z>-major and G'-major) ; and in Pfte. Son. op. loi, second move-
ment (March). — The Relative of the Subdom. in Beethoven, Pfte. Son.
op. 10, No. 2, second movement (/-minor and I>^-m&}or); also op. 14, No. i;
and op. 27, No. i, second movement of each.

See also : Chopin, Polonaises, No. 2, No. 3, No. 4.

^ Other possibilities of key-relation between the Prin. and

Subord. Songs are exhibited in the following :

Beethoven. Symphon y No. 7, third movement (Mediant-Dominant
relation, — F-msLJor and Z>-major,^he^ediant of the key of the Prin. Song
is the Dominant of the "Trio"). The same relation obtains in Haydn,
Symphony No. 3 (Peters ed.), third movement. — In Haydn, Symphony No. 2,
third movement, the keys arie Z> and jffji-major (Tonic-Mediant relation).-^*
The same in Chopin, Polonaise No. 6. — Brahms, op. 116, No. 6 (Dominant-
Relative key). Schubert, Impromptu op. 90, No. 2 (i&[>-major and ^-minor).

(c) The Tempo of th6 Subord. Song is very often, perhaps
usually, a little more tranquil than that of the Prin. Song ; though
it may be the reverse of this, and, sometimes, it remains unchanged.
This is dictated by the consideration of contrast^ which prevails,
mildly^ in this respect also. See :

Beethoven, Symphony No. 4, Menuetto (** Trio'' — un poco meno Allegro)\
Sym. No. 2» third movement (** Trio" — Presto meno assat). — Brahms, op. 116,
^TTb. 3 (Subord. Song— «« poco meno Allegro). — Chopin, Nocturne 13 {poco
piii lento).

Mendelssohn, Pfte. Son. op. 6, second movement (** Trio " — piu vivace),

Grieg, op. 38, No. 8 {piit mosso, ma tranquillo) .

In Beethoven, Pfte. Son. op. 26, second movement, — as in the majority of
cases, — no tempo-mark is given for the "Trio''; but it is manifestly more
tranquil in character than the Prin. Song, and will almost certainly suggest
a slight relaxation of speed, — to a thoughtful and susceptible performer.

(d) The Form of the '*Trio" is generally Three-Part; occa-
sionaTly only Two-Part. The One-Part form is very uncommon,
and can be chosen only when the Subord. Song is very distinct
in character, and when its Prin. Song contains at least two Parts,
—compare par. 11 83. See:

Beethoven, Pfte. Son. op. 2, No. i, Menuetto {^^ Trio* ^ 3-Part Song-form) ;
op. 10, No. 2, second movement (Subord. Song in 3-Part form, with modified
repetitions).

Beethoven, Pfte. Son. op. 26, second and third movements (** Trio*' in
each case 2-Part form); the same in op. 27, No. i, second movement, and
op. 27, No. 2, second movement.

Chopin, Mazurka 10 (Prin. Song 3-Part; Subord. Song only One-Part
form, but distinct); Mazurkas 45 and 46 (** Trio'* One-Part form); Mazurka 51

Par. 120. THE "DA CAPO. 215

("Trio** one large Part). — Schumann, Waldscenen^ op. 82, No. 7 (Prin.
SongTwo-Part form with Codetta; Subord. Song, one Part). — Bekthovkn»
Bagatelle^ op. 33, No. 4 (Subord. Song, one Part); Pfte. Son. op. 10, No. 3,
Menuetto (ditto). — Brahms, op. 116, No. 6, and op. 117, No. i, the same.

(e) While the Subord. Song, like the Principal one, is
expeCtecTEo terminate with a complete perfect cadence, — as implied
in par. ii6, — it is by no means uncommon to introduce a brief
Retransition after the Subord. Song, serving to lead smoothly
back into the recurrence of the Prin. Song, or '* Da capo." Comp.
par. 90c. It is more necessary than the Transition into the
Subord. Song (par. ii8c), because, the form having once been
expanded by the addition of a Subord. Song, the * * Da capo ' ' is
inevitable.

The Retransition may be independent of the Subord. Song, —

following after the complete cadence of the latter; or it may be

evolved by dissolution and modification of the cadence-member.

The details may be apprehended from the following examples :

Beethoven, Pfte. Son. op. 7, third movement, end of ** Minore " (two
measures of Re-transition); op. 10, No. 2, second movement (6 meas. of
independent Re-transition); op. 26, Scherzo {\ meas. of Re-transition); op. 2,
No. 3, Scherzo (the ** Trio" is in 3-Part form; Parts II and III are repeated,
and the latter is so modified,— dissolved, — at its end, as to lead away from its
own Tonic, into the Dominant of the Prin. Song, which follows as ** Da
capo"); op. 10, No. 3, Menuetto (similarly, the **Trio" ends on the Dom. of
the Prin. Song); op. 14, No. i, second movement (the same). — Beethoven,
Symp hony No. 3, and No. 5, third movement jiL-eack, — Haydn, Symphony
No. 2 (Peters edTJT'and No. 3, third Imovement of each. — Brahms, op. 116, No. 7
(elaborate independent Re-transition, after the Subord. Song); op. 117, No. 3
(the same).

The '*Da capo."

120. The recurrence of the Prin. Song, or "the so-called
'*Da capo," after the Subordinate Song, is, as a rule, literal.
When such is the case, it being unnecessary to write out the entire
Prin. Song again, its reproduction is merely indicated by the
words da capo (i. e., '*from the beginning"), or simply the letters
D. C ; or dal segno (i.e., '* from the sign " — ^ — ) in case the first
few tones or measures are so involved in the Re-transition as to
be excepted from the recurrence.

It is a rule, however, — with no other foundation than tradition,
— that such repetitions of the Parts as may have occurred at first,
are to be omitted in the *'Da capo"; hence the directions often
encountered : D, C, ma senza ripetizione.

2l6 THE *'DA capo." Par. 130:

See: Beethoven, Pfte. Son. op. 2, No. i, third movement, end of
"Trio"; op. 10, No. 3, Menueito, end of "Trio"; op. 22, Mcnuetto ; and
op. 26, SckerzOf end of "Trio." — Chopin, Polonaise I; Mazurka 6, 12, etc.

The ** Da capo," as a literal reproduction of the Prin. Song,
as implied above, the species which characterizes the ge nuine^
Song-form with **Trio."

But slight modifications or variations of the **Da capo," as
long as they remain thoroughly unessential^ are permissible. In
such cases the recurrence of the Prin. Song (Menuetto, Scherzo >
>or whatever it be) is, of course, written out.

N. B. It must be strictly borne in mind, that, as the modifi*

cations of the **Da capo" become more and more elaborate, the

design diverges in the same ratio from the specifically homophonic

Compound Song-form^ and approaches the spirit and detail of the

^ Rondo form.

■\

See Beethoven, Pfte. Son. op. 31, No. 3, Menuetto ; the "Da capo" is
written out, but not k tone is altered, excepting the final chord. Writing out
the literal reproduction of the Prin. Song appears whimsical, but it is often
involved by the Coda, and may be done purely for the convenience of the
player.

Chopin, Mazurka No. 23 (literal); No. 47 (literal, excepting repetitions);
Impromptu^ op. 29 (-^[^-major), literal; Impromptu, op. 66 (cjf-minor), literal,
excepting introduction. — Schubert, Momens musicals, op. 94, Nos. i and 4.

The '* Da capo " is slightly variated in Chopin, Nocturnes No. 11, No. 10,
No. I. — Brahms, op. 116, No. 3; op. 117, No. i; op. 117, No. 3 (repetitions
also omitted); op. 119, No. 2.

Somewhat more elaborate variation occurs in the ** Da capo " of Chopin,
^Nocturne No. 13; Beethoven, Pfte. Son. op. 27, No. i, second movement ^
and Beethoven, Bagatelle op. 33, No. 4.

In Chopin, Mazurka No. 36, the first 14 measures of the ** Da capo'' are
transposed (a half-step lower).

Al terations of the design of the Prin. Song, in the '*Da capo,'*
affect the genuineness of this class of forms more seriously than
simple variation of detail, because they are of a more essential
nature. The most natural, justifiable and common practice, is to
abbreviate the **Da capo" ;— especially by reducing the recurrence
of the 3-Part (or 2-Part) Song-form, to its First Part alone. Such
a contracted *' Da capo" may, subsequently, be extended, or
otherwise modified. See N. B. above.

See Beethoven, Bagatelle op. 33, No. 2. — Also, Symphony^
_ Scherzo.

Par. 121. THK CODA. 21*J

Chopin, Mazurkas 17 and 31 ; Mazurka 25 (Prin. Song, broad 2-Part
form, with complete repetition; **Da capo" abbreviated to First Part, slightly
expanded). Polonaise No. 4; Polonaise No. 6 (Prin. Song, 3-Part form, broad
First Part; "Trio," 2-Part form. Part I repeated. Part II dissolved into elabo-
rate Re-transition ; ** Da capo *' contracted to one-half of Part I). Impromptu
in Gt>-major, op. 51, (**Da capo*' abbreviated, and modified by extension);
the same in Nocturnes 3 and 14, Prelude, op. 28, No. 15, and Mazurka 38.

Brahms, op. 116, No. 6; op. 118, No. 4. ('*Da capo'* abbreviated and
modified.)

An extension of the ** Da capo " takes place in :

Chopin, Mazurkas 32 and 36; and Brahms, op. 116, No. 7.

The Coda.

121. The addition of a Coda to the entire Compound form
is not only possible, but desirable and necessary, apparently in
proportion to the importance attached to the **Da capo," and the
extent of its elaboration. Hence, while a Coda very rarely follows
the simple forms of the Song with ' * Trio ' ' as employed in older
Dances (with literal D. C), it is rarely, if ever, absent after a " Da
capo" that has been modified, and often it is quite extensive and
self-assertive. See par. 98a.

The Coda may be derived from any anterior motive or member ;
or it may (rarely) introduce new motives. And it may conform in
style and melodic contents either to the Prin. Song or the Subord.
Song (chiefly the former) ; or, possibly, to both in turn. See :

Beethoven, Pfte. Sonatas, op. 2, No. i; op. 2, No. 2; op. 7, etc., third
movement of each, — no Coda,

Beethoven, Pfte. Son. op. 2, No. 3, Scherzo (Coda derived from Prin.
Song); Pfte. Son. op. 14, No. i. Allegretto (Coda taken from Subord. Song,
* * Maggiore ") ; Pfte. Son. op. 26, third movement ("Funeral march"), last
7 measures; Pfte. Son. op. 27, No. i, second movement, last 12 measures
(Coda simply an extension of the **Da capo"). — Chopin, Mazurka No. 32
(elaborate Coda) ; No. 38 (Coda from Part II of Subord. Song). — Schubert,
Impromptu^ op. 90, No. 2 (Coda from **Trio"); Mendelssohn, Symphony
No. 4(** Italian," op. 90), third movement, — Coda utilizes ingeniously a motive
of the "Trio." — Brahms, op. 116, No. 6, last 7 measures (from Subord.
Song); op. 118, No. 4, last 23 measures; op. 119, No. 2, last 5 measures (from
Subord. Song).

MISCELLANEOUS EXAMPLES OF THE SONG-PORM WITH ONE ** TRIO.

>>

Schubert, Pfte. Sonatas, Nos. i, 2, 4, 6, 8, 9, 10, — third movement of each.
— Impromptu, op. 142, No. 2.

2l8 THE CODA. Par. 122.

Mozart, Symphonies (Litolff ed.), Nos. 2, 3, 4, 5, 6, 11, 12, 13, third
movement of each.

Brahms, Pfte. Son. op. i. Scherzo ; Pfte. Son. op. 2, Scherzo ; Pfte. Son.
op. 5, Scherzo ; Pfte. Balladen, op. 10, No. i ; Intermezzo, op. 76, No. 7 (Prin.
Song, One-Part form) ; Symphony No. 3, third movement

The String-quartets of Haydn, Mozart, Beethoven and other classic
writers, Menuetto or Scherzo (as a rule, the third movement of each).

1 22. A confusing trait of resemblance between the Song
with one '*Trio" and the simple 5 -Part Song-form, is exhibited
in certain diminutive examples of the former, where the Subord.
Song is only One-Part form, and the ** Da capo " abbreviated.

This is seen in Chopin, Mazurka 13 (Prin. Song 3-Part form; Subord.
Song only one Part, but very distinct in character and mode; **Da capo"
contracted to one of the original three Parts) . Precisely the same conditions
prevail in Grieg, Lyric Pieces, op. 12, No. 3. — These are both examples of the
5-Part Song-form, stage 5, cited in par. io6e; but they are extreme specimens,
owing to the unusual independence and diversity of the Fourth Part^ — ^which,
for this reason, assumes the rank and characteristics of a "Trio." Were the
**Da capo" not reduced to one Part, the identity of such a Subord. Song
would be beyond question.

In Chopin, Mazurka 42, again only five Parts are represented, — the Prin.
Song being in t^vo Parts, and the Subord. Song in one ; but here the ** Da
capo '"* is not abbreviated (save by omission of repetitions).

EXERCISE 35.

A. An example of the Song-form with one •* Trio," in major ; both Songs
in 3-Part form; **Da capo '* literal ; no Transition or Re-transition. — ^Apply
par. 136.

B. An example of the same form, in minor; design of Songs optional;
" Da capo " literal. — Apply par. 134, or par. 135.

C. Same form; design of Songs optional; Re- transition ; **Da capo"
unessentially modified. — Apply par. 136.

D. Same form; **Da capo" abbreviated; Coda. — Apply par. 135. See
par. 137.

tt r,^,>T^O "

Par. 184. THE SONG-FORM WITH TWO "TRIOS. 219

CHAPTER XIX.

EXTENSIONS OF THE "SONG WITH TRIO."

1 23. The simplest method of extension is, here again, that
of Repetition. It is applied, not infrequently, to the Subord.
Song and the **Da capo" together^ precisely as in the 3-Part
Song-form with repeated Second and Third Parts. See par. lo^a.

The repetition may be exact or modified, — generally the
former, though it is natural to abbreviate the last '''•Da capo^^'' —
more rarely the first one. A Coda may follow.

See Beethoven, Symphony No. 4, third movement (Prin. Song —
**Minuetto*' — 3-Part form with all repetitions, and Codetta; ** Trio '* 3-Part
form; Re-transition; ist "Da capo" literal, excepting repetitions; **Trio"
again, exactly as before; second **Da capo'' contracted to Third Part; Coda

measures). — In Beethove n. Symphony No. y . thi rd move ment, the same
repetition occurs, in the original score ; a Coda is added, reverting very briefly
to the motive of the Subord. Song. (In some modern editions, and modern
performances, this extensive repetition is omitted.)— Beethoven, Bagatelle
op* 33> No. 7, (both ** Songs" in concise 2-Part form, with repetitions; both
**Da capos" complete, but variated; Coda). — Beethoven, Pfte. Son. op. 54,
first movement (the Subord. Song is at first an extended Period, with transposed
and enlarged reproduction, and Re-transition ; when it recurs, after the first
** Da capo,'' it is abbreviated to the length of a simple Period; the two **Da
capos " are variated, in progressive degrees; Coda).

In Haydn, Pfte. Son. No. 7 (Cotta ed.). Finale^ the Subord. Song and
** Da capo" are repeated with considerable, but probably unessential, modifi-
cation (Prin. Song, 2-Part form; Subord. Song, 3-Part; no Coda). Pfte.
Son. No. 8, Scherzando (very similar; brief Coda). Pfte. Son. No. 12, first
movement, the same.

In Schumann, Symphony No. 4 (^/-minor, op. 120), Scherzo^ the **Trio"
recurs, similarly, after the ** Da capo," but is linked with the Transition which
leads into the Finale ; in other words, the second ** Da capo " is omitted.

The Song-form with Two *' Trios.

>>

1 24. When such an enlargement of the form is contem-
plated, it is, however, much better and more customary to avoid
the monotony attendant upon so extensive a repetition, by invent-
ing a new {second) ^^Trio^^' instead of the recurrence of the first
one. This design corresponds to that of the fully developed
5-Part Song-form (par. 106^), of which it is a broader exposition.
Compare par. 117, second clause.

220 THE GROUP OF SONG-FORMS. Par. 125.

The two ' ' Trios ' ' should stand in quite marked contrast with
each other ; one of them generally maintaining closer agreement,
in character and style, with the Prin. Song, while the other
diverges more emphatically. Consequently, it is not unusual for
the Subord. Songs to differ from each other in Time^ as well as in
tempo and key ; though this is more commonly the case in modern
examples.

After each Subord. Song, the Prin. Song recurs, as *' Da
capo," sometimes literally (excepting the repetitions), but more
frequently abbreviated, or otherwise modified. A Coda may be
added.

See JoHANN LuDWiG Krebs, Partita II, ^*Menuets*^ (three, numbered
I, II, III, — No. I recurring as ** Da capo'* after each of the others). [To be
found in '' I/es Maltres du clavecin '' (Litolff ed.). Vol. I, p. 73.]

Mozart, Symphony No. 8 (Litolff ed.), fourth movement, — Menuetto
with two ** Trios" (not to be confounded with the second movement of the
Symphony) .

Mendelssohn, Wedding- March from music to ** Midsummer-Night's
Dream."

Schumann, Symphony No. i. Scherzo (Prin. Song 3-Part form ; **Trio I "
very divergent in character, time, and tempo; broad form, with modified
repetitions; first "Da capo" without repetitions; **Trio II" resembles Prin.
Song; 3-Part form, with Codetta from Second Part; second "Da capo*'
abbreviated; Coda contains reminiscences of ** Trio I"). — Symphony No. 2,
Scherzo. — Pfte. Quintet, op. 44, Scherzo. — Pfte. Quartet, op. 47, Scherzo.—'
Pfte. Trio, No. 3, op. no, third movement.

Brahms, Pfte. Scherzo, op. 4. — Symphony No. 2, third movement (Prin.
Song J time. Allegretto, 3-Part form; First Subord. Song, same key, \ time,
Presto f 3-Part form; first ** Da capo" abbreviated and modified; Second
Subord. Song, | time. Presto ; second **Da capo" complete, but transposed
during the first Part; Coda). In this unique example, the two "Trios"
represent, essentially, two similar extreme Variations of the Prin. Song".

The Group of Song-forms.

1 25. The process of group-formation, beginning with the
Phrase-group (par. 57) and passing on through the Group of
Periods (par. 59), the Large Phrase-group (par. 104) and the Group
of Parts (par. 114, 115), culminates, in the homophonic forms, in
the Group of entire Song-forms. It is distinguished from the
regular Compound forms, explained above, by the absence, or
irregular disposition, of the *'Z>« capo,'''' whereby the condition
of '* Return, after Departure," is not fully or correctly satisfied
(par. 8 1 a). See par. 69c.

Par. 125. THE GROUP OF SONG-FORMS. 231

See Beethoven, Pfte. Son. op. io6, Scherzo (First, — or Principal, — Song,
2 Parts, each repeated ; Second Song, ditto ; Third Song, change of time and
tempo, 3 Parts ; Fourth Song like First, variated ; Coda. The form is quasi
"Song with tivo Trios in succession^"*).

Chopin, Polonaise No. V, the same; (the Third Song, — or 2nd ** Trio,'' —
is in Two-Part form, very broad, and with complete transposed reproduction) .

Brahms, Pfte. Ballade op. lo, No. 2, the same design. Rhapsody, op. 119,
No. 4, (design : Prin. Song— *' Trio I"— "Trio 11"— ** Trio I "— Prin. Song,
Coda). Pfte. Ballade op. 10, No. 4, (design: Prin. Song— *' Trio I "—Prin.
Song — ** Trio II" — Coda, consisting in a portion of the First Subord. Song^
interwoven with the first motive of the Prin. Song. Thie final **Da capo" is
omitted).

Schumann, Nachtstiicke^ op. 23, No. 2, (design: Prin. Song — **Trio V'
7-Partform— Re-transition— Prin. Song, abbreviated— "Trio II "—"Trio I"—
Prin. Song). — Symphony No. 3 Scherzo^ (design: Prin. Song — Song II —
Song III — Song IV, transposition of Prin. Song — Song V — Song VI, " Da
capo" of Prin. Song — Coda).

Chopin, Waltz No. i (Group of 5 Songs, V like I).

These represent the last degree of enlargement, the broadest
proportions attainable within the sphere of the Homophonic
Forms.

EXERCISE 36.

A. An example of the Song with two ** Trios," in major, first ** Da capo "
abbreviated. — ^Apply par. 133 {Scherzo),

B. The same, in minor. Apply par. 136.

322 CONVENTIONAL STYLES OF COMPOSITION. Pai.tOBb,

DIVISION FOUR.

CONVENTIONAL STYLES OF COMPOSITION

1 26. An extremely large proportion of all music written,
belongs to the Homophonic domain of composition. The entire
range of musical products within this domain may be approxi-
mately divided into ^hree general classes or styles of composition,
distinguished by the respective predominance of one of the three
essential factors of the art : Melody, Harmony and Rhythm.
(Carefully review par. 97, recollecting, however, that the distinc-
tions there defined constitute a different classification from the one
under present consideration.)

(a) The first, or Lyric, class, in which the element of
Melody predominates, is characterized chiefly by the Song (with
or without words), and embraces also the Air or Aria, Lied,
Canzone, Cavatina, Idyll, Barcarolle or Gondellied, Romanza,
Reverie, Nocturne, Serenade, *' Melody," Chanson, Lyric Piece,
Ballade, Elegy, Berceuse or Cradle-song, Pastorale, — and many
other compositions of similar character but with more or less
arbitrary and specific titles. All of these may be either vocal or
instrumental in conception and setting. To this class belong also
the sacred Hymn, Psalm-tune, Chorale, Chant, the simpler church
Anthems, the secular Terzetto, Quartet, Chorus, Glee, Madrigal,
etc., all of vocal conception.

(b) The second, or ]&tude-, class, in which the most promi-
nent element is that of the Harmony, is characterized chiefly by
the Etude, and includes also the various kinds of Studies and
Exercises ; the Toccata ; certain varieties of the Prelude ; generally
also the Caprice or Capriccio, Scherzo, Scherzando, Impromptu,
Intermezzo, — though the character of these latter is indefinite and
variable. The compositions of this class are almost invariably
instrumental in conception and setting.

Par. 127. CONVENTIONAL STYLES OF COMPOSITION. 223

(c) The third, or Dance-, class, distinguished by the prev-
alence of the element of Rhythm, embraces all Dances, old and
modern, and the March. Conspicuous among these are the Min-
uet, Gavotte, Courante, Sarabande, Gigue or Jig, Waltz, Polka,
Galop, Mazurka, Tarantella, Saltarello, Siciliano, Landler, Bolero,
Quadrilles, Polonaise, Marches of varied character (Wedding-M.,
Funeral-M., Festival-M., etc.), and many other species of a more
or less kindred nature. This class of compositions is very gener-
ally, though not always, instrumental in conception and setting.

(d) It must be remembered that this is only an approximate classification,
and that, while some of the conventional styles enumerated have become
distinctly typical, others again are so indefinite and arbitrary as to defy exact
classification. Hence, it is often impracticable and unwise to apply these
class-distinctions rigorously. It is perfectly just to impart a conspicuously
melodic character to a Dance-form, as long as the rhythmic element maintains
its preeminence ; or to an fitude, — though, in the latter case, it would transfer
the composition to the lyric class, and the reasons for calling it an " fitude'*
would have to be sought in other, technical, peculiarities of the piece. This
accounts for the apparently careless or erroneous choice of title in certain
examples; many a Nocturne or Waltz, so-called, belongs properly to the
£tude-class, — and xnce versa.

As a rule, however, the student should exercise judgment in naming
his compositions. It is best never to use descriptive titles (almost utterly
senseless when applied to music — e.g. *' Among the roses,'* — "Shower of
pearls," etc.), but only such conventional titles as the above, which indicate
the general or typical musical characteristics of the piece. A very important
lesson is conveyed to the young composer by Brahms, in his almost exclusive
use of the terms Capriccio, Intermezzo, Fantasie, for his pianoforte-pieces;
and in the choice of the simple tempo designations: "Allegretto, *' "Andante,'*
etc., adopted as sole title by many serious composers.

127- A description of each of these conventional species
of composition may be found in standard dictionaries of music,
and in such books as the "Musical Forms" by E. Pauer. All
that need be added, here, is a purely technical definition of such
structural traits as concern the present student, who is expected
to write an example of each, or at least of the most important, of
the typical varieties noted.

224

THE SONG, WITH WORDS.

Par. 128b.

CHAPTER XX.

THE LYRIC CLASS.

I. The Song, with Words.

1 28d.. As concerns the first, and apparently difficult, consid-
eration of choosing the text for a Song, the student will find the
richest and most trustworthy fund of suitable words in the volumes
of Songs already written and published. From these he can
choose, studiously ignoring the musical setting before him. Or he
may take a Psalm, or some other Bible passage ; or may adopt
a poem from books of standard poetry, though this is the least
advisable course for the beginner to pursue.

The text should first be thoroughly memorized, and mentally
repeated, until it begins to suggest consistent musical setting. This
will facilitate the choice of general characteristics, i. e., between
duple and triple Time, major and minor mode, brisk or deliberate
tempo, bold or graceful style.

(b) In the setting, the rhythmic distinctions of the musical
meter must coincide quite accurately with the prosody of the text.
Important ivords and accented syllables should be placed against
the accented^ or higher^ or longer^ tones ; and lower, shorter or
unaccented tones should accompany unemphatic words and syl-
lables.

This important rule operates, however, mainly by comparison: an un-
accented word or syllable may be set to an accented tone, if the accent is
subordinate ; or to a higher tone, if it occupies a li^/tt beat. For illustration :

♦1) «2) *3)

lOl.

m

*

^m

'*Gol - den vis - ions'

'Gol'deu

*i

"Ool - den vis - ions'*

*4)

♦5)

i=t

-#-5-

-#-S-

T— #-T-

-«»--

Gol - den vis- ions" **Ool - den vis -ions of the past,"

i

l=r:t

^i

"o/ the past"

*7)

-^-

-9-

-^-

' damn

- %ng.

i

Par. 128e.

THE SONG, WITH WORDS.

225

*i) Good, because the accent at c is subordinate to that at g^ and at the
following h. — *2) Doubtful, because of the emphasis attached to a compara-
tively very high tone. — *3) Better, because the high tone is unaccented. —
*4) Eccentric. — *5) Objectionable. — *6) Better. — *7) Such an accent as this,
upon a light final syllable, is permissible at any cadence.

Furthermore, this rule must always be applied with sufficient latitude to
ensure perfectly unconstrained melodious conduct of the vocal part. The
words are to be regarded and treated, constantly, as the subordinate element
of a Song.

(c) In writing for the human voice, it is necessary to regard
the average compass of the part in question :

Bass.

* A {.±)

Tekob.

102.

m

{■*•)

-^ — »

^

-^■.

Alto.

SorKAXO.

^m

x:!*!

i

♦ (feA)

I

(■«^)

(■■^)

The range of Baritone is between Bass and Tenor; that of
Mezzo-Soprano, between Alto and Soprano.

Care must be taken to avoid maintaining, persistently, either
a comparatively high, or low, range of the chosen voice. This
consideration will influence the choice of key.

(d) The student should endeavor to reflect the character of the
text in the mood of the music. He must determine the dynamic
and declamatory design, according to the dramatic undulations of
the text ; but must avoid exaggerated minuteness in this respect.
The effort to ''illustrate" each single suggestive word leads to
unevenness of structure, and hampers the essentially musical con-
ception. This is the gravest error that can attach itself to a vocal
composition ; for a Song must he firsts and always^ good^ melodious,
and self- sufficient music,

(e) In the NOTATION of the vocal part, the notes which
accompany separate words or syllables must be, detached; two or
more notes to the same word or syllable must be connected, either
by beams, ties, or slurs :

226

THE SONG, WITH WORDS.

Par. 128f .

Ij^

loa a^^

t

X

X

Brahms.

^=^

X

■zt

**Shr(yud'ed in sad - mss" "aS is dark,"

- ful sun - beams,'*

(f ) The FORM of the Song will depend largely upon that of
the text, which may demand almost any of the forms explained
in chapters VI to XVII of this book, from the large (or repeated)
Period, up to the 3- or 5-Part form, or Group of Parts. But
preference should be given to the Two- and Three- Part Song-
forms, and an effort be made to adapt the text to such a design,
before commencing the composition.

When the text is divided (or divisible) into stanzas, the
composition may be strophic, i.e., music may be set to the first
stanza only, and simply reproduced, with more or less essential
change, for the following ones. See :

Schubert, **Mllller-Lieder" (op. 25), Nos. i, 7. 8, 9, 10, 13, etc.

Beethoven, "An die feme Geliebte'* (op. 98), sections a, ^, c, </, e;
(section f is in 3-Part form ; the final section is a Coda, compounded out of
the first section). This cycle of Songs ranks among the most beautiful and
impressive in all musical literature. The student is to examine it thought-
fully. Its design is, of course, the Group of Song-forms (par. 125).

Schumann, "Liederkreis'* (op. 24), Nos. 4, 7.

Or the composition maybe progressive, i.e., the words may
be set to music consecutively, throughout, — with, probably, a da
capo at the end, as in some of the Group-forms. In this case, the
3-Part or 5-Part Song, or the Group of Parts, — possibly Song with
'* Trio," — may be chosen as structural design. See :

Schubert, "MUller-Lieder,'' No. 2,— Song with** Trio"; (Prin. Song,
3-Part Period; Subord. Song, 3-Part Song-form; **Da capo," a compound
of both foregoing Songs; Coda). No. 3, — Group of four Parts ; (I, a Period ;
II, a Phrase, repeated; III, an extended Phrase; IV, two repeated Phrases).
No. 4, — quasi 3 strophes. No. 5, — Song with ** Trio."

Schumann, **Liederkreis" (op. 24), Nos. i, 2 (almost strophic). No. 3
(Three-Part form), No. 5 (Fiire-Part form) ; ** The two Grenadiers," op. 49,
No. I (partly strophic).

Par. 128g. THE SONG, WITH WORDS. 227

(gC) The primary objects of the instrumental accompaniment
to a Song are, first, to support^ and secondly, to complement ^ the
vocal part. It should not disturb or overpower the latter, and
therefore the danger of too much accompaniment must be recognized
and avoided.

The simple rhythmic enunciation of the chords, as in Schubert,
** Haidenroslein'* (op. 3, No. 3) is antiquated, but often very appropriate.
The method adopted by Schumann in the ** Lotosblume*' (op. 25, No. 7), in
which the vocal melody is from time to time reinforced by the pianoforte, is
more artistic.

But the student must guard against the impression that the vocal part
should be duplicated throughout in the accompaniment (as in Schumann,
op. 25, Nos. I, 2, 9, etc.). This tends to obscure, more than to support, the
vocal part; and, while it is often necessary, it is better, as a rule, for the
accompaniment to limit itself to general harmonic figuration, — which is at
once less obtrusive, and more individualized. See Schumann, op. 25, Nos. 3,
II, 25. Schubert, *' MUller-Lieder^' (op. 25), Nos. i, 2, 3, 4, 5, 9, 11, etc. —
A duplication of the vocal melody, as found in Schubert, " Winterreise "
(op. 89), No. 15; or in Schumann, op. 25, No. 10; and op. 27, No. 4, — is very
effective.

But it is permissible, and common, to assign to the instru-
mental accompaniment the task of reflecting and emphasizing the
poetic, epic, or dramatic contents of the text ; and, in so doing, it
will become more or less characteristic, and assume a degree of
independence, which, at times^ may even exceed that of the vocal
part itself.

See Schubert, ** MtiUer-Lieder " No. 3 (Bass part). No. 12; "Winter-
reise," Nos. 2, 4, 9, 17. Schumann, op. 24, No. 6; op. 27, No. 3; op. 30,
No. i; "Lowenbraut," op. 31, No. i; ** Frauenliebe und -leben," op. 42,
No. 6; "Dichterliebe," op. 48, Nos. 4, 6, 9, 13, 16.

This will give rise most naturally to the prelude, interludes

and postlude in the accompaniment (during the pauses of the

vocal part) ; though the necessity of occasional interludes^ — ^not too

frequent or long, — both for the relief of the singer and the benefit

of the structural design, is sufficiently imperative, in itself.

See Schubert, ** Winterreise," Nos. 2, 6, 14, 19, 24 (especially).
Schumann, op. 25, No. 3; op. 24, Nos. 5, 6 (elaborate postludes). " Dich-
terliebe," Nos. 5, 6, 9, 10, 11, 12, 15, 16 (elaborate postludes).

Additional examples for reference and study :

Other Songs of Schubert, Beethoven and Schumann, not referred to
above.

The Songs of Mendelssohn, Brahms, Robert Franz; Rubinstein,
Grieg, Jensen, Taubert; Song-albums (Schirmer ed.).

228 THE SONG WITHOUT WORDS, ETC. Par. 130a.

2. The instrumental Duo.

1 28. The instruments most frequently chosen for the duo
are the Pianoforte and Violin, or Pfte. and Violoncello.

The rules for the conception and treatment of this style of
music conform in general to those of the Song (excepting those
bearing upon the text), but with such additional liberty, or
modifications, as are conditioned by

The wider compass,

greater technical facilities and resources, and

specific peculiarities of the ** solo "-instrument.

The absence of words leaves the composer without a certain
quality of melodic stimulus and suggestion, but, at the same time,
at liberty to develop his musical conception and purpose without
embarrassment. Some theoretical guides to his natural musical
impulses are recorded in par. 94, 95, 96, 97, which are to be
reviewed. See also par. 137.

The student, if unfamiliar with the Violin and Violoncello,
should not undertake to write for them until he has consulted some
expert performer, and obtained from him sufficient information
concerning the tone, specific tone-effects, compass, and the various
technical characteristics of the instrument. The same law applies
to the use of the Flute, Clarinet, Cornet, Horn, etc.

The species of composition for the instrumental duo is to be
selected from the list given in the next paragraph (or par. 126 a).

Examples for reference : *

Violin and Pianoforte : Raff, Cavatine. — Svendsen, Romance, op. 26. —
H. WiENiAWSKi, L6gende, op. 17. — F. Ries, Romance from Suite II;
Gondoliera from Suite III. — Max Bruch, Swedish Dances, op. 63 (essentially
lyric); Romanze. — Jadassohn, Serenade op. 108^, Nos. i, 2, 3, 4. — Spohr,
Barcarolle. — Bazzini, ]Sl6gie. — Joachim, Romanza. — DvoiiAK, Notturno,
op. 40. — Ernst, fil6gie. — Vieuxtemps, Reverie.

Violoncello and Pianoforte: Mendelssohn, Song without Words for
^ Cello (Z?-major, op. 109). — DvoKAk, ** Waldesruhe." — Popper, op. 3, Nos. 2,
6; Nocturne, op. 22. — Goltermann, Cantilena in E. — H. Sitt, Romanza and
Serenata. — Davidoff, Romance sans paroles, op. 23; Lied, op. 16, No. 2.

3. The Song without Words, etc.

1 30a.. The conception of this class of instrumental com-
positions, most commonly set for the Pianoforte, or as duo

*See Preface, section II.

^

Par. 181a. THE HYMN, ANTHEM, GLEE, ETC. 239

(par. 129), is invariably the same as that of the vocal Song, viz.:
a continuous, coherent, tuneful melodic thread, — cantilena^ — as
distinctly predominating element; with a more or less character-
istic and elaborate equipment in the accompanying parts. Each
specific variety of conventional style embraced under this heading,
— Romanza, Nocturne, Barcarolle, etc. (par. 126a), — will call
forth its own peculiar modifications, in the execution of this aim ;
but the ruling principle is identical in them all. These modifica-
tions, dictated by the special character and purpose of the chosen
variety, may be inferred from the self-explaining titles,, and need
no exposition here. See Grove's dictionary, or Baker's ** Dic-
tionary of Musical Terms." Any form may be employed, from
2-Part Song up to Song with one **Trio." See par. 137.
Examples, from pianoforte literature, for reference :

Chopin, Nocturnes (especially Nos. i, 2, 4, 5, 7, 8, 11, 13, 15, 16, 17, 18, 19).
— Field, Nocturnes (Nos. i, 2, 4, 5, 6, 7, 9, 10, 11, 13, 14, 16, 18). — Mendels-
sohn, ** Songs without Words" (especially Nos. i, 6, 7, 12, i8, 19, 22, 25, 29,
30» 31* 36, 37, 40, 43, 46). — Schubert, Impromptu op. 90, No. 3; Momens
tnusicals, op. 94, No. 2. — Beethoven, Pfte. Son. op. 27, No. i, Adagio ; op.
27, No. 2, first movement; op. 79, Andante; op. 109, third movement, Theme;
op. no, third movement, Arioso dolente, — Schumann, ** Kinderscenen " (op.
15), Nos. I, 4, 5, 7, 12; ** Fantasiestiicke'* (op. 12), Nos. 1,3; Pfte. Son. op. 22,
Andantino; Romanzas (op. 28), Nos. i, 2 ; op. 82, No. 9. — W. Bargiel, fil6gie,
op. 31, No. 2. — Dvoft-AK, Silhouettes, op. 8, Nos. 6, 10, 11. — Gade, Idyls, op.
34, Nos. 2, 4. — Grieg, op. 12, Nos. i and 7 ; op. 38, Nos. i, 3, 6; op. 43, No. 6;
op. 47, Nos. 3, 7; op. 54, No. 4; op. 62, No. 5. — Liszt, 3 Notturnos (**Liebes-
traume "); Gondoliera and Canzone from ** Venezia e Napoli " ; Consolations.
— MoszKowsKi, op. 31, No. I ; op. 36, No. 2.— ^aff, 2 £l6gies, op. 149. —
Rubinstein, 2 melodies, op. 3 ; Barcarolle, op. 50, No. 3 (^-minor) ; also
Barcarolles in rt-minor,y-minor, and G-major. — Paderewski, M6lodie, op. 8.
No. 3. — St.-Sa£ns, Chanson, op. 72, No. 5. — Tschaikowsky, Chanson, op. 40,
No. 2. ; Romance, op. 51, No. 5.

(b) The Ballade belongs, properly, also to this class, but its
character is indefinite and variable. It is usually more elaborate,
longer, and more dramatic than the Romanza, and constructed
in Group-form. It implies a more or less ideal narrative.

See Chopin, 4 Ballades. — Brahms, op. 10, Nos. i, 2, 4. — Reinecke,
Ballade, op. 20.

4. The Hymn, Anthem, Glee, etc.

131a.. It may appear that these species of composition,
commonly known as * 'Part-Songs," belong to the second class
(par. 1263), characterized by predominance of the Harmony; and.

230 THE HYMN, ANTHEM, GLEE, ETC. Par. 131e.

indeed, this fallacious view has been often confirmed by writers not
sufficiently scrupulous in regard to artistic distinctions. But it is
nevertheless strictly true that, in those examples of the * 'Part-
Song " which fall within the domain of the homophonic forms, the
lyric element must prevail. One of the parts, usually the Soprano,
must consist in a distinctly melodious, coherent and continuous
cantilena^ which the other parts chiefly serve to accompany and
support.

(b) For the Sacred Hymn, Chorale, Anthem, etc., words
may be chosen from a church hymn-book, the Psalms and other
parts of the Bible, or from the ritual of any denominational service.
The use of Latin sentences from the Roman Catholic liturgy is
strongly commended.

(c) The ensemble most commonly adopted for the setting,
is the mixed quartet (Soprano, Alto, Tenor, Bass) ; but the female
trio (more rarely female quartet), the male trio or quartet, the
mixed trio, and the duet (the association of any two voices), are all
equally practicable and often very effective. In any case, except-
ing perhaps the duet, the *' parts" may be rendered by solo singers,
or by a chorus ; and either with instrumental accompaniment (Pfte.
or Organ), or without — (''« cappella ").

(d) The choice of secular^ instead of sacred, text, will
influence the character of the musical conception, but none of
the above details. The dignity and seriousness which should
distinguish all music designed for religious use and association, is
more or less thoroughly supplanted, in secular music, by such
brightness, gaiety, grace, dramatic fervor, pathos or brilliancy, as
befits the character of the words selected.

(e) The structural design will depend largely upon the text.
For the Hymn, the Double- period form, or simple 2-Part Song-
form, is best. For the Anthem, or the secular *' Part-Song, " the
3-Part form should be used, if possible; though the "Group of
Parts" is often more convenient, and affords excellent opportunity
for characteristic and interesting formal designs, in which, for
instance, the several Parts of the Group may be differently treated
(as solo, duet, quartet or chorus, perhaps in different varieties of
time, tempo and character). The conditions of the instrumental

Par. 132. THE ^TUDE, OR STUDY. 23 1

accompaniment are similar to those explained in connection with
the **Song" (par. 128^), though, as a rule, less elaborate and
independent. For more specific definitions, see Grove's dic-
tionary.

See also: Mendelssohn, female Terzetto from Elijah; Duet from
St. Paul (No. 31); also Chorales (Nos. 3, 9, 16), and Choruses (Nos. 26, 33)
from St. Paul.

The vocal duets of Mendelssohn, Rubinstein, Schumann, and
Brahms.

Brahms, "Ave Maria"; 13th Psalm; Song from Ossian^s Fingal.

Anthem-books for mixed voices (Schirmer ed.). — Reference may also be
made to a few specimens of the " Part-Songs," sacred and secular, for various
ensembles of male, female, or mixed voices, contained in great number in the
8vo collections of the edition of G. Schirmer ; and similar collections of other
publishers.

CHAPTER XXI.

THE ETUDE-CLASS.
I. The 6tudk, or Study.

1 32. In the genuine representatives of this class of com-
position, the melodic element (as sustained cantilena)^ though
never totally absent, is so vague, imperfect and fragmentary, or so
disguised and obscured, as to recede into the background, while
the chords, i. e., the Harmonies, both individually and collectively,
stand out in proportionately greater prominence.

In the ^TUDE proper, the harmonies are not, as a rule, thus
prominent in unbroken bulk, but in some figurated form (as shown
in Ex. 4 and context). For this figuration of the harmony, a
motive is adopted with a view to some technical purpose ; hence
the titles : Etude, Study, or Exercise. But, for the simple reason
that the aim of a ** study" is not always a purely technical one,
these species may, without inconsistency, sometimes assume a dis-
tinctly lyric character. See par. 126^.

The design of the Etude is usually one of the Song- forms, —
possibly, though rarely, with **Trio."

See Chopin, fitudes, op. 10 (especially Nos. i, 2, 5, 7, 8, 10, 11, 12);
Etudes, op. 25 (especially Nos. 2, 3, 4, 5, 6, 8, 9, 10, 11, 12). Op. 10, Nos. 3, 6,
9, and op. 25 No. 7, are lyric. — Ad. Hensklt, fitudes, op. 5, No. i, (both lyric
and 6tude-cla6s), Nos. 2, 3, 7, 9, 12 (10 and 11, lyric). — See also a /<?w Etudes

232 THE TOCCATA, CAPRICCIO, SCHERZO, ETC. Par. 183d.

of Clementi, Cramer and Czerny. — Liszt, three Concert-etudes; Etudes
transcendentales. — Mendelssohn, Etudes op. 104, Nos. i, 3. — Moscheles,
Studies, op. 70. — Raff, ** La fileuse," op. 157, No. 2. — Schumann, op. 3. —
Rubinstein, op. 23, and op. 81. — Tschaikowsky, op. 40, No. i. —

2. The Toccata, Capriccio, Scherzo, etc.

1 33a. When the motive or figure, upon which the *' figura-
tion" or dissolution of the harmony is to be based, assumes more
of a thematic character, and thus enters more essentially into the
structure of the composition than is the case with the chiefly
technical figures of the Study, the style is usually designated
Toccata (especially when the figure is small), or Caprice, and
Prelude. Review par. 126^.

In the latter species, and also in the Impromptu, Intermezzo,
etc., the harmony frequently appears in unbroken, or but partially
broken, bulk.

(b) The Scherzo was originally nearly identical with the
*' Caprice." Later it was substituted for the Minuet in. the
Symphony, Sonata, and Quartet, and adopted the time (J) of that
Dance. Finally, it superseded the latter, appearing in duple as
well as triple time, and often as an independent piece, without
reference to the Symphony.

(c) In all of these species of composition the element of
Melody, it must be remembered, is necessarily always present
in a more or less apparent and assertive degree^ but neither as
conspicuously, nor as continuously, as in the lyric class. Certain
sections of an Etude, Toccata, etc., or of a Dance-form (e. g.,
the **Trio"), may be purely lyric; and, as already .stated, the
distinctions of style, especially among the etude-species, are quite
vague, — often scarcely perceptible. Review par. 126^.

(d) These species are usually, though by no means always,
written in the homophonic forms. In this case, the designs
chosen are the Song-forms, — rarely with **Trio," excepting in the
Scherzo, which sometimes has two •* Trios."

Examples for reference :

Toccata-species, and Prelude: Bach, Well-tempered Clavichord, Vol. I,
Preludes i, 2, 5, 6, 15, 21; Vol. II, Preludes 3 (ist section), 12, 15; Partita
No. I, **Gigue.^' — Bargiel, fitude and Toccata, op. 45. — Schubert, Momens
musicals, op. 94, Nos. 4, 5. — Schumann, Arabesque, op. 18; op. 21, No. 2. —

Par. IM. OLD DANCE-SPKCIES. 233

Heller, Praeludien, op. 81. — Chopin, Preludes, op. 28; op. 45. — St.-Sa£ns,
Toccata, op. 72, No. 3. — Mendelssohn, ** Songs without Words," Nos. 8, 24,
34, 38. — Rubinstein, Preludes, op. 24, Nos.- 1, 4.

Caprice, Impromptu, etc. : Chopin, Impromptus, op. 29, op. 51, op. 66. —
Gade, Aquarelles, op. 57, Nos. i, 3, 5; Fantasies, op. 41, Nos. 2, 3. — Grieg,
op. 43, Nos. I, 4; op. 62, No. 4. — MoszKOWSKi, op. 7, No. 2,; op. 36, No. 6.
— Schubert, Impromptu, op. 90, No. 4. — Rubinstein, Caprices, op. 21. —
TscHAiKOWSKY, Capriccio, op. 8. — Schumann, Intermezzi, op. 4; op. 12,
Nos. 2, 5, 6, 7, 8.

Scherzo: Beethoven, Pfte. Son. op. 2, No. 2, third movement; op. 26,
second movement; op. ^8, third movement. — Chopin, Pfte. Son. op. 35, second
movement; Scherzos, op. 20, 31, 39, 54. — Grieg, op. 54, No. 5. — Mendelssohn,
" Songs without Words," No. 45; op. 16, No. 2. — Schumann, Pfte. Son. op.
14, second movement; op. 22, third movement; op. 12, No. 4; op. 21, Nos. i,
3, 6; op. 26, No. 3; op. 32, No. i. — Schubert, Pfte. Sonatas Nos. i and 9
(Peters ed.) third movement of each. — Hummel, Pfte. Son. op. 106, second
movement. — Brahms, op. 4.

CHAPTER XXII.

THE DANCE-CLASS.
I. Old Dance-spkcies.

1 34. Very many of the older dances have fallen into disuse,
and the corresponding musical species are therefore seldom written
expressly for the purpose of dancing. But the old types, with their
respective rhythmic peculiarities, are nevertheless often adopted,
and invested with the characteristics and charms of modern musical
setting.

For the list of old dance-forms, and their several details, refer-
ence must be made to Grove's dictionary, or some standard book
upon '* Dances," old and modern. See par. 126^, and par. 127 ; and
examine, carefully, the following examples of these dance-species
as found, in their original condition (in older writings), and in the
idealized expositions of more modern composers :

Allemande, Courante, Bourr^, Sarabande, Passepied : Bach, English
Suites; French Suites; Partitas. — Les MaItres du Clavecin (Litolff ed.),
Vol. I, pages 68, 69, 70, 82-86, 117-119, 166, 167, 189; Vol. II, pages 16, 46, 47,
109, no, 174 (Tambourtn), 178, 221 {Galliardo),

Qavotte: Bach, Engl. Suites, Nos. 3, 6; French Suites, Nos. 5, 6 (Ex. 75
of this book). — Les MaItres du Clavecin, Vol. I, pages 50, 191; Vol. II,
p. 28.

234 THE MARCH. Par. 136.

Qigue; Les MaItrks du, Clavecin, Vol. I, p. 58, 74, 93, 132, 178, 192;
Vol. II, p. 48, 50, III, 142, 144, 189.

Minuet: Bach, Engl. Suite, No. 4; French Suites, Nos. i, 2, 3, 6;
Partitas, Nos. i, 4. — Les MaItres du Clavecin, Vol. I, p. 128, 143; Vol. II,
p. 150, 187, 196, 200, 208. — Haydn, Pfte. Son. No. i (Cotta ed.), second move-
ment; Symphonies and String-quartets. — Mozart, Pfte. Son. No. 9 (Cotta ed.),
second movement; No. 12, tHird movement; Symphonies and String-quartets.
— Beethoven, Pfte. Son. op. 10, No. 3, third movement; op. 22, ditto; op. 31,
No. 3, ditto. — Schubert, Pfte. Sonatas, Nos. 4, 8, third movement of each.
— Mendelssohn, Son. op. 6, second movement. — Grieg, op. 57, No. i. —

PADEREWSKI, Op. I4, No. I.

Further: Bargiel, Suite of old Dances, op. 7. — Rubinstein, ditto, op.
38. — MoszKOWSKi, Bourr6e, op. 38, No. i. — Raff, Tambourin, op. 204, No. 6.
—Silas, Gavotte in e-minor.

2. Modern Dance-spkcies.

135. The most prominent place among modern dances with
musical setting is assigned to the Waltz, Mazurka, Polonaise,
Polka, Tarantella, Quadrille, and a few others of national, rather
than universal, importance. Their musical exposition is sometimes
brief and simple ; but more commonly they are idealized, or elabo-
rated into ** Concert-pieces " of considerable length and freedom
of form, without neglect, however, of the distinctive rhythmic
peculiarities of the respective species. Information concerning the
latter may be obtained from the authorities already cited, and from
careful inspection of the following examples :

Chopin, Waltzes; Mazurkas (especially Nos. i, 5, 6, 12, 13, 14, 17, 18, 20,
22, 24, 25, 26, 32, 36, 41, 47) ; Polonaises (especially Nos. i, 2, 3, 4, 6, 8, 9, 10) ;
Tarantella, op. 43; Bolero, op. 19. — Johann Strauss, difew Waltzes and other
Dances (Schirmer ed.). — Brahms, Waltzes for 4 hands, op. 39; Hungarian
Dances, for 2 or 4 hands. — DvorAk, Waltzes, op. 54; Slavonian Dances, for
2 or 4 hands, op. 46, op. 72. — Grieg, op. 12, Nos. 2, 5, 6; op. 38, Nos. 4, 5 ;
op. 47, Nos. I, 6; op. 62, No. i. — Liszt, Polonaise in E major; Tarantella
(** Venezia e Napoli**). — Moszkowski, Waltzes, op. 8; op. 46, No. i. —
Rubinstein, " Le Bal,'* op. 14. — St.-Sa£ns, Waltzes, op. 72, No. 4; op. 104.
— Tschaikowsky, Mazurka, op. 9, No. 3. — Weber, ^'Invitation to the Dance '*
(Waltz).

3. The March.

1 36. Though not a "Dance," in the specific sense of the
term, the March belongs in the foremost rank of that class of
musical compositions in which marked rhythm is the ruling trait
and purpose.